Andreas Geiger (MPI Tübingen) | Philip Lenz (KIT) | Christoph Stiller (KIT) | Raquel Urtasun (University of Toronto)

Object Detection Evaluation 2012


The object detection and object orientation estimation benchmark consists of 7481 training images and 7518 test images, comprising a total of 80.256 labeled objects. All images are color and saved as png. For evaluation, we compute precision-recall curves for object detection and orientation-similarity-recall curves for joint object detection and orientation estimation. In the latter case not only the object 2D bounding box has to be located correctly, but also the orientation estimate in bird's eye view is evaluated. To rank the methods we compute average precision and average orientation similiarity. We require that all methods use the same parameter set for all test pairs. Our development kit provides details about the data format as well as MATLAB / C++ utility functions for reading and writing the label files.

We evaluate object detection performance using the PASCAL criteria and object detection and orientation estimation performance using the measure discussed in our CVPR 2012 publication. For cars we require an overlap of 70%, while for pedestrians and cyclists we require an overlap of 50% for a detection. Detections in don't care areas or detections which are smaller than the minimum size do not count as false positive. Difficulties are defined as follows:

  • Easy: Min. bounding box height: 40 Px, Max. occlusion level: Fully visible, Max. truncation: 15 %
  • Moderate: Min. bounding box height: 25 Px, Max. occlusion level: Partly occluded, Max. truncation: 30 %
  • Hard: Min. bounding box height: 25 Px, Max. occlusion level: Difficult to see, Max. truncation: 50 %

All methods are ranked based on the moderately difficult results. Note that for the hard evaluation ~2 % of the provided bounding boxes have not been recognized by humans, thereby upper bounding recall at 98 %. Hence, the hard evaluation is only given for reference.
Note 1: On 25.04.2017, we have fixed a bug in the object detection evaluation script. As of now, the submitted detections are filtered based on the min. bounding box height for the respective category which we have been done before only for the ground truth detections, thus leading to false positives for the category "Easy" when bounding boxes of height 25-39 Px were submitted (and to false positives for all categories if bounding boxes smaller than 25 Px were submitted). We like to thank Amy Wu, Matt Wilder, Pekka Jänis and Philippe Vandermersch for their feedback. The last leaderboards right before the changes can be found here!

Note 2: On 08.10.2019, we have followed the suggestions of the Mapillary team in their paper Disentangling Monocular 3D Object Detection and use 40 recall positions instead of the 11 recall positions proposed in the original Pascal VOC benchmark. This results in a more fair comparison of the results, please check their paper. The last leaderboards right before this change can be found here: Object Detection Evaluation, 3D Object Detection Evaluation, Bird's Eye View Evaluation.
Important Policy Update: As more and more non-published work and re-implementations of existing work is submitted to KITTI, we have established a new policy: from now on, only submissions with significant novelty that are leading to a peer-reviewed paper in a conference or journal are allowed. Minor modifications of existing algorithms or student research projects are not allowed. Such work must be evaluated on a split of the training set. To ensure that our policy is adopted, new users must detail their status, describe their work and specify the targeted venue during registration. Furthermore, we will regularly delete all entries that are 6 months old but are still anonymous or do not have a paper associated with them. For conferences, 6 month is enough to determine if a paper has been accepted and to add the bibliography information. For longer review cycles, you need to resubmit your results.
Additional information used by the methods
  • Stereo: Method uses left and right (stereo) images
  • Flow: Method uses optical flow (2 temporally adjacent images)
  • Multiview: Method uses more than 2 temporally adjacent images
  • Laser Points: Method uses point clouds from Velodyne laser scanner
  • Additional training data: Use of additional data sources for training (see details)

Car


Method Setting Code Moderate Easy Hard Runtime Environment
1 CLOCs_PVCas 95.96 % 96.76 % 91.08 % 0.1 s 1 core @ 2.5 Ghz (Python)
S. Pang, D. Morris and H. Radha: CLOCs: Camera-LiDAR Object Candidates Fusion for 3D Object Detection . 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2020.
2 BorderAtt 95.88 % 96.68 % 90.89 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
3 PVGNet 95.80 % 96.87 % 93.05 % 0.05 s 1 core @ >3.5 Ghz (C/C++)
4 ADLAB 95.69 % 96.69 % 90.81 % 0.08 s 1 core @ >3.5 Ghz (C/C++)
5 HUAWEI Octopus 95.50 % 96.30 % 92.81 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
6 SPANet 95.46 % 96.54 % 90.47 % 0.06 s 1 core @ 2.5 Ghz (C/C++)
7 PC-CNN-V2
This method makes use of Velodyne laser scans.
 95.20 % 96.06 % 89.37 % 0.5 s GPU @ 2.5 Ghz (Matlab + C/C++)
X. Du, M. Ang, S. Karaman and D. Rus: A General Pipeline for 3D Detection of Vehicles. 2018 IEEE International Conference on Robotics and Automation (ICRA) 2018.
8 F-PointNet
This method makes use of Velodyne laser scans.
 code 95.17 % 95.85 % 85.42 % 0.17 s GPU @ 3.0 Ghz (Python)
C. Qi, W. Liu, C. Wu, H. Su and L. Guibas: Frustum PointNets for 3D Object Detection from RGB-D Data. arXiv preprint arXiv:1711.08488 2017.
9 SA-SSD code 95.16 % 97.92 % 90.15 % 0.04 s 1 core @ 2.5 Ghz (Python)
C. He, H. Zeng, J. Huang, X. Hua and L. Zhang: Structure Aware Single-stage 3D Object Detection from Point Cloud. CVPR 2020.
10 Voxel R-CNN 95.11 % 96.49 % 92.45 % 0.04 s GPU @ 3.0 Ghz (C/C++)
11 3DSSD code 95.10 % 97.69 % 92.18 % 0.04 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Yang, Y. Sun, S. Liu and J. Jia: 3DSSD: Point-based 3D Single Stage Object Detector. CVPR 2020.
12 MVRA + I-FRCNN+ 94.98 % 95.87 % 82.52 % 0.18 s GPU @ 2.5 Ghz (Python)
H. Choi, H. Kang and Y. Hyun: Multi-View Reprojection Architecture for Orientation Estimation. The IEEE International Conference on Computer Vision (ICCV) Workshops 2019.
13 MMLab PV-RCNN
This method makes use of Velodyne laser scans.
 code 94.70 % 98.17 % 92.04 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
S. Shi, C. Guo, L. Jiang, Z. Wang, J. Shi, X. Wang and H. Li: PV-RCNN: Point-Voxel Feature Set Abstraction for 3D Object Detection. CVPR 2020.
14 IC-SECOND 94.69 % 95.79 % 91.89 % 0.06 s 1 core @ 2.5 Ghz (C/C++)
15 PC-RGNN 94.68 % 95.80 % 92.20 % 0.1 s GPU @ 2.5 Ghz (Python + C/C++)
16 D3D 94.66 % 95.43 % 89.72 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
17 Deformable PV-RCNN
This method makes use of Velodyne laser scans.
 code 94.64 % 95.86 % 92.10 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
P. Bhattacharyya and K. Czarnecki: Deformable PV-RCNN: Improving 3D Object Detection with Learned Deformations. ECCV 2020 Perception for Autonomous Driving Workshop.
18 nonet 94.62 % 95.86 % 91.86 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
19 CN 94.60 % 97.86 % 89.81 % 0.04 s GPU @ 2.5 Ghz (Python + C/C++)
20 HyBrid Feature Det 94.59 % 95.76 % 92.18 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
21 ReFineNet 94.59 % 95.75 % 92.12 % 0.08 s 1 core @ 2.5 Ghz (Python)
22 MGACNet 94.57 % 95.35 % 91.77 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
23 RangeRCNN-LV 94.51 % 95.93 % 92.07 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
24 OneCoLab SicNet V2 94.48 % 95.65 % 92.11 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
25 TuSimple code 94.47 % 95.12 % 86.45 % 1.6 s GPU @ 2.5 Ghz (Python + C/C++)
F. Yang, W. Choi and Y. Lin: Exploit all the layers: Fast and accurate cnn object detector with scale dependent pooling and cascaded rejection classifiers. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2016.
K. He, X. Zhang, S. Ren and J. Sun: Deep residual learning for image recognition. Proceedings of the IEEE conference on computer vision and pattern recognition 2016.
26 AP-RCNN 94.47 % 96.13 % 90.15 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
27 EPNet code 94.44 % 96.15 % 89.99 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
T. Huang, Z. Liu, X. Chen and X. Bai: EPNet: Enhancing Point Features with Image Semantics for 3D Object Detection. ECCV 2020.
28 SERCNN
This method makes use of Velodyne laser scans.
 94.42 % 96.33 % 89.96 % 0.1 s 1 core @ 2.5 Ghz (Python)
D. Zhou, J. Fang, X. Song, L. Liu, J. Yin, Y. Dai, H. Li and R. Yang: Joint 3D Instance Segmentation and Object Detection for Autonomous Driving. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition 2020.
29 IC-PVRCNN 94.38 % 95.89 % 91.90 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
30 CVRS_PF 94.37 % 95.56 % 91.43 % 0.09 s 1 core @ 2.5 Ghz (C/C++)
31 CVIS-DF3D_v2 94.33 % 95.70 % 91.72 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
32 UberATG-MMF
This method makes use of Velodyne laser scans.
 94.25 % 97.41 % 89.87 % 0.08 s GPU @ 2.5 Ghz (Python)
M. Liang*, B. Yang*, Y. Chen, R. Hu and R. Urtasun: Multi-Task Multi-Sensor Fusion for 3D Object Detection. CVPR 2019.
33 SVGA-Net
This method makes use of Velodyne laser scans.
 94.24 % 95.86 % 91.80 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
34 Baseline of CA RCNN 94.23 % 95.84 % 91.80 % 0.1 s GPU @ 2.5 Ghz (Python)
35 CVIS-DF3D 94.23 % 95.84 % 91.80 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
36 tbd code 94.21 % 95.68 % 91.49 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
37 CA-RCNN 94.20 % 95.81 % 91.53 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
38 HR-faster-rcnn 94.14 % 95.41 % 86.88 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
39 RangeRCNN
This method makes use of Velodyne laser scans.
 94.03 % 95.48 % 91.74 % 0.06 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Liang, M. Zhang, Z. Zhang, X. Zhao and S. Pu: RangeRCNN: Towards Fast and Accurate 3D Object Detection with Range Image Representation. arXiv preprint arXiv:2009.00206 2020.
40 OAP 93.93 % 96.85 % 86.37 % 0.06 s 1 core @ 2.5 Ghz (C/C++)
41 HRI-MSP-L
This method makes use of Velodyne laser scans.
 93.92 % 95.51 % 91.42 % 0.07 s 1 core @ 2.5 Ghz (C/C++)
42 AF_V1 93.87 % 94.45 % 86.37 % 0.1 s 1 core @ 2.5 Ghz (Python)
43 Associate-3Ddet_v2 93.77 % 96.83 % 88.57 % 0.04 s 1 core @ 2.5 Ghz (Python)
44 Patches - EMP
This method makes use of Velodyne laser scans.
 93.75 % 97.91 % 90.56 % 0.5 s GPU @ 2.5 Ghz (Python)
J. Lehner, A. Mitterecker, T. Adler, M. Hofmarcher, B. Nessler and S. Hochreiter: Patch Refinement: Localized 3D Object Detection. arXiv preprint arXiv:1910.04093 2019.
45 CIA-SSD
This method makes use of Velodyne laser scans.
 93.72 % 96.87 % 86.20 % 0.03 s 1 core @ 2.5 Ghz (Python + C/C++)
46 VAL 93.71 % 96.92 % 83.76 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
47 deprecated 93.68 % 96.92 % 86.15 % deprecated deprecated
48 Noah CV Lab - SSL 93.65 % 94.02 % 86.02 % 0.1 s GPU @ 2.5 Ghz (Python)
49 CBi-GNN 93.60 % 98.89 % 88.47 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
50 MVX-Net++ 93.58 % 96.41 % 88.51 % 0.15 s 1 core @ 2.5 Ghz (C/C++)
51 CLOCs_PointCas 93.55 % 96.69 % 86.16 % 0.1 s GPU @ 2.5 Ghz (Python)
52 MonoPair 93.55 % 96.61 % 83.55 % 0.06 s GPU @ 2.5 Ghz (Python + C/C++)
Y. Chen, L. Tai, K. Sun and M. Li: MonoPair: Monocular 3D Object Detection Using Pairwise Spatial Relationships. The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2020.
53 CIA-SSD v2
This method makes use of Velodyne laser scans.
 93.52 % 96.63 % 88.21 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
54 Deep MANTA 93.50 % 98.89 % 83.21 % 0.7 s GPU @ 2.5 Ghz (Python + C/C++)
F. Chabot, M. Chaouch, J. Rabarisoa, C. Teulière and T. Chateau: Deep MANTA: A Coarse-to-fine Many-Task Network for joint 2D and 3D vehicle analysis from monocular image. CVPR 2017.
55 Point-GNN
This method makes use of Velodyne laser scans.
 code 93.50 % 96.58 % 88.35 % 0.6 s GPU @ 2.5 Ghz (Python)
W. Shi and R. Rajkumar: Point-GNN: Graph Neural Network for 3D Object Detection in a Point Cloud. CVPR 2020.
56 PP-3D 93.50 % 96.58 % 88.35 % 0.1 s 1 core @ 2.5 Ghz (Python)
57 FCY
This method makes use of Velodyne laser scans.
 93.49 % 96.74 % 88.39 % 0.02 s GPU @ 2.5 Ghz (Python)
58 CJJ 93.48 % 96.68 % 90.63 % 0.04 s 1 core @ 2.5 Ghz (C/C++)
59 AIMC-RUC 93.47 % 96.75 % 88.35 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
60 PointRes
This method makes use of Velodyne laser scans.
 93.47 % 96.69 % 90.46 % 0.013 s 1 core @ 2.5 Ghz (Python + C/C++)
61 dgist_multiDetNet 93.46 % 94.99 % 85.46 % 0.08 s GPU Titanx Pascal (Python)
62 scssd-normal(0.3) 93.45 % 96.72 % 88.31 % 0.05 s GPU @ 2.5 Ghz (Python)
P. An, J. Liang, J. Ma, K. Yu and B. Fang: SCSSD for multi sensor fusion 3d object detection method. Submitted to Information Science 2020.
63 Cas-SSD 93.41 % 96.73 % 88.30 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
64 RRC code 93.40 % 95.68 % 87.37 % 3.6 s GPU @ 2.5 Ghz (C/C++)
J. Ren, X. Chen, J. Liu, W. Sun, J. Pang, Q. Yan, Y. Tai and L. Xu: Accurate Single Stage Detector Using Recurrent Rolling Convolution. CVPR 2017.
65 KNN-GCNN 93.39 % 96.19 % 88.17 % 0.4 s 1 core @ 2.5 Ghz (Python + C/C++)
66 F-3DNet 93.38 % 96.51 % 88.32 % 0.5 s GPU @ 2.5 Ghz (Python)
67 HR-Cascade-RCNN 93.37 % 95.74 % 87.44 % 0.3 s 1 core @ 2.5 Ghz (C/C++)
68 3D-CVF at SPA
This method makes use of Velodyne laser scans.
 93.36 % 96.78 % 86.11 % 0.06 s 1 core @ 2.5 Ghz (C/C++)
J. Yoo, Y. Kim, J. Kim and J. Choi: 3D-CVF: Generating Joint Camera and LiDAR Features Using Cross-View Spatial Feature Fusion for 3D Object Detection. ECCV 2020.
69 ISF 93.36 % 96.64 % 90.52 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
70 FLID 93.35 % 95.90 % 85.69 % 0.04 s GPU @ 2.5 Ghz (Python)
71 ISF-v2 93.34 % 96.73 % 90.54 % 0.04 s 1 core @ 2.5 Ghz (Python)
72 scssd-normal(0.4) 93.31 % 96.59 % 88.23 % 0.05 s 1 core @ 2.5 Ghz (Python)
P. An, J. Liang, J. Ma, K. Yu and B. Fang: SCSSD for multi sensor fusion 3d object detection method. Submitted to Information Science 2020.
73 STD code 93.22 % 96.14 % 90.53 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Yang, Y. Sun, S. Liu, X. Shen and J. Jia: STD: Sparse-to-Dense 3D Object Detector for Point Cloud. ICCV 2019.
74 SARPNET 93.21 % 96.07 % 88.09 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
Y. Ye, H. Chen, C. Zhang, X. Hao and Z. Zhang: SARPNET: Shape Attention Regional Proposal Network for LiDAR-based 3D Object Detection. Neurocomputing 2019.
75 RoIFusion code 93.19 % 96.29 % 88.14 % 0.22 s 1 core @ 3.0 Ghz (Python)
76 Fast Point R-CNN
This method makes use of Velodyne laser scans.
 93.18 % 96.13 % 87.68 % 0.06 s GPU @ 2.5 Ghz (Python + C/C++)
Y. Chen, S. Liu, X. Shen and J. Jia: Fast Point R-CNN. Proceedings of the IEEE international conference on computer vision (ICCV) 2019.
77 sensekitti code 93.17 % 94.79 % 84.38 % 4.5 s GPU @ 2.5 Ghz (Python + C/C++)
B. Yang, J. Yan, Z. Lei and S. Li: Craft Objects from Images. CVPR 2016.
78 RethinkDet3D 93.14 % 96.16 % 88.17 % 0.15 s 1 core @ 2.5 Ghz (Python)
79 Discrete-PointDet 93.14 % 96.36 % 87.82 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
80 SJTU-HW 93.11 % 96.30 % 82.21 % 0.85s GPU @ 1.5 Ghz (Python + C/C++)
S. Zhang, X. Zhao, L. Fang, F. Haiping and S. Haitao: LED: LOCALIZATION-QUALITY ESTIMATION EMBEDDED DETECTOR. IEEE International Conference on Image Processing 2018.
L. Fang, X. Zhao and S. Zhang: Small-objectness sensitive detection based on shifted single shot detector. Multimedia Tools and Applications 2018.
81 BLPNet_V2 93.11 % 96.07 % 88.06 % 0.04 s 1 core @ 2.5 Ghz (Python)
82 PVF-NET 93.08 % 96.03 % 88.04 % 0.1 s 1 core @ 2.5 Ghz (Python)
83 SerialR-FCN+SG-NMS 93.03 % 95.81 % 83.00 % 0.2 s 1 core @ 2.5 Ghz (Python)
84 NLK-ALL code 92.98 % 95.73 % 88.13 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
85 CLOCs_SecCas 92.95 % 95.43 % 89.21 % 0.1 s 1 core @ 2.5 Ghz (Python)
S. Pang, D. Morris and H. Radha: CLOCs: Camera-LiDAR Object Candidates Fusion for 3D Object Detection. 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2020.
86 cvMax 92.84 % 96.14 % 87.87 % 0.04 s GPU @ >3.5 Ghz (Python)
87 TBD 92.82 % 96.06 % 88.00 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
88 HotSpotNet 92.81 % 96.21 % 89.80 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
Q. Chen, L. Sun, Z. Wang, K. Jia and A. Yuille: object as hotspots. Proceedings of the European Conference on Computer Vision (ECCV) 2020.
89 deprecated 92.79 % 95.56 % 91.62 % 0.06 s 1 core @ 2.5 Ghz (C/C++)
90 deprecated 92.79 % 96.12 % 87.78 % 0.04 s GPU @ 2.5 Ghz (Python)
91 PointCSE 92.78 % 95.99 % 87.66 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
92 IGRP 92.78 % 96.28 % 87.81 % 0.18 s 1 core @ 2.5 Ghz (Python + C/C++)
93 Mono3CN 92.76 % 95.51 % 84.80 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
94 MuRF 92.74 % 95.74 % 87.64 % 0.05 s GPU @ 1.5 Ghz (Python + C/C++)
95 SegVoxelNet 92.73 % 96.00 % 87.60 % 0.04 s 1 core @ 2.5 Ghz (Python)
H. Yi, S. Shi, M. Ding, J. Sun, K. Xu, H. Zhou, Z. Wang, S. Li and G. Wang: SegVoxelNet: Exploring Semantic Context and Depth-aware Features for 3D Vehicle Detection from Point Cloud. ICRA 2020.
96 Patches
This method makes use of Velodyne laser scans.
 92.72 % 96.34 % 87.63 % 0.15 s GPU @ 2.0 Ghz
J. Lehner, A. Mitterecker, T. Adler, M. Hofmarcher, B. Nessler and S. Hochreiter: Patch Refinement: Localized 3D Object Detection. arXiv preprint arXiv:1910.04093 2019.
97 CenterNet3D 92.69 % 95.76 % 89.81 % 0.04 s GPU @ 1.5 Ghz (Python)
G. Wang, B. Tian, Y. Ai, T. Xu, L. Chen and D. Cao: CenterNet3D:An Anchor free Object Detector for Autonomous Driving. 2020.
98 Chovy 92.69 % 96.06 % 89.74 % 0.04 s GPU @ 2.5 Ghz (Python)
99 PPFNet code 92.68 % 96.32 % 87.66 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
100 R-GCN 92.67 % 96.19 % 87.66 % 0.16 s GPU @ 2.5 Ghz (Python)
J. Zarzar, S. Giancola and B. Ghanem: PointRGCN: Graph Convolution Networks for 3D Vehicles Detection Refinement. ArXiv 2019.
101 NLK-3D 92.67 % 95.44 % 87.72 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
102 PI-RCNN 92.66 % 96.17 % 87.68 % 0.1 s 1 core @ 2.5 Ghz (Python)
L. Xie, C. Xiang, Z. Yu, G. Xu, Z. Yang, D. Cai and X. He: PI-RCNN: An Efficient Multi-sensor 3D Object Detector with Point-based Attentive Cont-conv Fusion Module. AAAI 2020 : The Thirty-Fourth AAAI Conference on Artificial Intelligence 2020.
103 deprecated 92.60 % 96.20 % 89.60 % - -
104 deprecated 92.59 % 96.21 % 89.58 % 0.05 s GPU @ >3.5 Ghz (Python)
105 PointPainting
This method makes use of Velodyne laser scans.
 92.58 % 98.39 % 89.71 % 0.4 s GPU @ 2.5 Ghz (Python + C/C++)
S. Vora, A. Lang, B. Helou and O. Beijbom: PointPainting: Sequential Fusion for 3D Object Detection. CVPR 2020.
106 VICNet 92.56 % 95.56 % 87.40 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
107 DEFT 92.55 % 96.17 % 89.51 % 1 s GPU @ 2.5 Ghz (Python)
108 3D IoU-Net 92.47 % 96.31 % 87.67 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
J. Li, S. Luo, Z. Zhu, H. Dai, S. Krylov, Y. Ding and L. Shao: 3D IoU-Net: IoU Guided 3D Object Detector for Point Clouds. arXiv preprint arXiv:2004.04962 2020.
109 PPBA 92.46 % 95.22 % 87.53 % NA s GPU @ 2.5 Ghz (Python)
110 TBU 92.46 % 95.22 % 87.53 % NA s GPU @ 2.5 Ghz (Python)
111 VAR 92.46 % 95.11 % 89.68 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
112 Associate-3Ddet code 92.45 % 95.61 % 87.32 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
L. Du, X. Ye, X. Tan, J. Feng, Z. Xu, E. Ding and S. Wen: Associate-3Ddet: Perceptual-to-Conceptual Association for 3D Point Cloud Object Detection. The IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2020.
113 OneCoLab SicNet 92.37 % 95.57 % 89.79 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
114 Dccnet 92.34 % 96.00 % 86.85 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
115 LZnet 92.34 % 95.35 % 89.33 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
116 PointRGCN 92.33 % 97.51 % 87.07 % 0.26 s GPU @ V100 (Python)
J. Zarzar, S. Giancola and B. Ghanem: PointRGCN: Graph Convolution Networks for 3D Vehicles Detection Refinement. ArXiv 2019.
117 LZY_RCNN 92.28 % 93.58 % 89.76 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
118 F-ConvNet
This method makes use of Velodyne laser scans.
 code 92.19 % 95.85 % 80.09 % 0.47 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Wang and K. Jia: Frustum ConvNet: Sliding Frustums to Aggregate Local Point-Wise Features for Amodal 3D Object Detection. IROS 2019.
119 MDA 92.17 % 94.88 % 89.54 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
120 PFF3D
This method makes use of Velodyne laser scans.
 92.15 % 95.37 % 87.54 % 0.05 s GPU @ 3.0 Ghz (Python + C/C++)
121 yolo4 92.13 % 94.20 % 79.89 % 0.02 s 1 core @ 2.5 Ghz (Python)
122 TBD 92.12 % 93.48 % 89.56 % 0.05 s GPU @ 2.5 Ghz (Python)
123 PVNet 92.12 % 94.84 % 89.27 % 0,1 s 1 core @ 2.5 Ghz (Python)
124 SRDL
This method uses stereo information.
This method makes use of Velodyne laser scans.
 92.11 % 95.59 % 89.60 % 0.15 s GPU @ 2.5 Ghz (Python + C/C++)
125 IE-PointRCNN 92.08 % 96.01 % 87.05 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
126 PBASN code 92.07 % 95.51 % 87.04 % NA s GPU @ 2.5 Ghz (Python)
127 SDP+RPN 92.03 % 95.16 % 79.16 % 0.4 s GPU @ 2.5 Ghz (Python + C/C++)
F. Yang, W. Choi and Y. Lin: Exploit All the Layers: Fast and Accurate CNN Object Detector with Scale Dependent Pooling and Cascaded Rejection Classifiers. Proceedings of the IEEE International Conference on Computer Vision and Pattern Recognition 2016.
S. Ren, K. He, R. Girshick and J. Sun: Faster R-CNN: Towards real-time object detection with region proposal networks. Advances in Neural Information Processing Systems 2015.
128 AB3DMOT
This method makes use of Velodyne laser scans.
This is an online method (no batch processing).
 code 92.00 % 95.88 % 86.98 % 0.0047s 1 core @ 2.5 Ghz (Python)
X. Weng and K. Kitani: A Baseline for 3D Multi-Object Tracking. arXiv:1907.03961 2019.
129 DPointNet 92.00 % 95.01 % 86.86 % 0.09 s 1 core @ 2.5 Ghz (C/C++)
130 MMLab-PointRCNN
This method makes use of Velodyne laser scans.
 code 91.90 % 95.92 % 87.11 % 0.1 s GPU @ 2.5 Ghz (Python + C/C++)
S. Shi, X. Wang and H. Li: Pointrcnn: 3d object proposal generation and detection from point cloud. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2019.
131 MMLab-PartA^2
This method makes use of Velodyne laser scans.
 code 91.86 % 95.03 % 89.06 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
S. Shi, Z. Wang, J. Shi, X. Wang and H. Li: From Points to Parts: 3D Object Detection from Point Cloud with Part-aware and Part-aggregation Network. IEEE Transactions on Pattern Analysis and Machine Intelligence 2020.
132 Pointpillar_TV 91.82 % 94.82 % 88.57 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
133 epBRM
This method makes use of Velodyne laser scans.
 code 91.77 % 94.59 % 88.45 % 0.1 s GPU @ >3.5 Ghz (Python + C/C++)
K. Shin: Improving a Quality of 3D Object Detection by Spatial Transformation Mechanism. arXiv preprint arXiv:1910.04853 2019.
134 deprecated 91.76 % 96.53 % 83.90 % 0.05 s 1 core @ 2.5 Ghz (Python)
135 3DBN_2 91.75 % 95.34 % 89.12 % 0.12 s 1 core @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
136 C-GCN 91.73 % 95.64 % 86.37 % 0.147 s GPU @ V100 (Python)
J. Zarzar, S. Giancola and B. Ghanem: PointRGCN: Graph Convolution Networks for 3D Vehicles Detection Refinement. ArXiv 2019.
137 ITVD code 91.73 % 95.85 % 79.31 % 0.3 s GPU @ 2.5 Ghz (C/C++)
Y. Wei Liu: Improving Tiny Vehicle Detection in Complex Scenes. IEEE International Conference on Multimedia and Expo (ICME) 2018.
138 yolo4_5l 91.71 % 93.35 % 79.49 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
139 PiP 91.67 % 94.35 % 88.35 % 0.033 s 1 core @ 2.5 Ghz (Python)
140 SINet+ code 91.67 % 94.17 % 78.60 % 0.3 s TITAN X GPU
X. Hu, X. Xu, Y. Xiao, H. Chen, S. He, J. Qin and P. Heng: SINet: A Scale-insensitive Convolutional Neural Network for Fast Vehicle Detection. IEEE Transactions on Intelligent Transportation Systems 2019.
141 Faster RCNN + A 91.60 % 94.77 % 81.43 % 0.19 s GPU @ 2.5 Ghz (Python + C/C++)
142 Cascade MS-CNN code 91.60 % 94.26 % 78.84 % 0.25 s GPU @ 2.5 Ghz (C/C++)
Z. Cai and N. Vasconcelos: Cascade R-CNN: High Quality Object Detection and Instance Segmentation. arXiv preprint arXiv:1906.09756 2019.
Z. Cai, Q. Fan, R. Feris and N. Vasconcelos: A unified multi-scale deep convolutional neural network for fast object detection. European conference on computer vision 2016.
143 tt code 91.59 % 95.15 % 88.72 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
144 V3D 91.52 % 94.46 % 86.34 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
145 MAFF-Net(DAF-Pillar) 91.46 % 94.38 % 83.89 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
Z. Zhang, Z. Liang, M. Zhang, X. Zhao, Y. Ming, T. Wenming and S. Pu: MAFF-Net: Filter False Positive for 3D Vehicle Detection with Multi-modal Adaptive Feature Fusion. arXiv preprint arXiv:2009.10945 2020.
146 HRI-VoxelFPN 91.44 % 96.65 % 86.18 % 0.02 s GPU @ 2.5 Ghz (Python + C/C++)
H. Kuang, B. Wang, J. An, M. Zhang and Z. Zhang: Voxel-FPN:multi-scale voxel feature aggregation in 3D object detection from point clouds. sensors 2020.
147 RUC 91.40 % 95.02 % 88.41 % 0.12 s 1 core @ 2.5 Ghz (C/C++)
148 CU-PointRCNN 91.34 % 97.25 % 86.98 % 0.1 s GPU @ 1.5 Ghz (Python + C/C++)
149 deprecated 91.31 % 96.90 % 83.91 % 0.06 s GPU @ >3.5 Ghz (Python)
150 Faster RCNN + G 91.28 % 94.34 % 81.02 % 1.1 s GPU @ 2.5 Ghz (Python)
151 Faster RCNN + Gr + A 91.25 % 94.09 % 81.25 % 1.29 s GPU @ 2.5 Ghz (Python)
152 OACV 91.21 % 94.23 % 83.07 % 0.23 s GPU @ 2.5 Ghz (Python)
153 CentrNet-FG 91.21 % 94.05 % 88.45 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
154 PointPillars
This method makes use of Velodyne laser scans.
 code 91.19 % 94.00 % 88.17 % 16 ms 1080ti GPU and Intel i7 CPU
A. Lang, S. Vora, H. Caesar, L. Zhou, J. Yang and O. Beijbom: PointPillars: Fast Encoders for Object Detection from Point Clouds. CVPR 2019.
155 Faster RCNN + A 91.19 % 94.43 % 80.99 % 0.19 s GPU @ 2.5 Ghz (Python)
156 LTN 91.18 % 94.68 % 81.51 % 0.4 s GPU @ >3.5 Ghz (Python)
T. Wang, X. He, Y. Cai and G. Xiao: Learning a Layout Transfer Network for Context Aware Object Detection. IEEE Transactions on Intelligent Transportation Systems 2019.
157 autonet 91.17 % 93.70 % 88.10 % 0.12 s 1 core @ 2.5 Ghz (C/C++)
158 WS3D
This method makes use of Velodyne laser scans.
 91.15 % 95.13 % 86.52 % 0.1 s GPU @ 2.5 Ghz (Python)
Q. Meng, W. Wang, T. Zhou, J. Shen, L. Van Gool and D. Dai: Weakly Supervised 3D Object Detection from Lidar Point Cloud. 2020.
159 PointPiallars_SECA 91.12 % 93.66 % 87.94 % 0.06 s 1 core @ 2.5 Ghz (C/C++)
160 EPENet 91.11 % 94.31 % 88.02 % 0.04 s 1 core @ 2.5 Ghz (C/C++)
161 anonymous 91.08 % 96.57 % 82.86 % 1 s 1 core @ 2.5 Ghz (C/C++)
162 SSL-RTM3D 91.07 % 96.44 % 81.19 % 0.03 s 1 core @ 2.5 Ghz (Python)
163 FII-CenterNet 91.03 % 94.48 % 83.00 % 0.09 s GPU @ 2.5 Ghz (Python)
164 Aston-EAS 91.02 % 93.91 % 77.93 % 0.24 s GPU @ 2.5 Ghz (Python + C/C++)
J. Wei, J. He, Y. Zhou, K. Chen, Z. Tang and Z. Xiong: Enhanced Object Detection With Deep Convolutional Neural Networks for Advanced Driving Assistance. IEEE Transactions on Intelligent Transportation Systems 2019.
165 ARPNET 90.99 % 94.00 % 83.49 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
Y. Ye, C. Zhang and X. Hao: ARPNET: attention region proposal network for 3D object detection. Science China Information Sciences 2019.
166 Bit 90.96 % 93.84 % 87.47 % 0.11 s 1 core @ 2.5 Ghz (C/C++)
167 JSU-NET 90.90 % 96.41 % 80.67 % 0.1 s 1 core @ 2.5 Ghz (Python)
168 GAFM 90.90 % 96.46 % 80.70 % 0.5 s 1 core @ 2.5 Ghz (Python)
169 PatchNet code 90.87 % 93.82 % 79.62 % 0.4 s 1 core @ 2.5 Ghz (C/C++)
X. Ma, S. Liu, Z. Xia, H. Zhang, X. Zeng and W. Ouyang: Rethinking Pseudo-LiDAR Representation. Proceedings of the European Conference on Computer Vision (ECCV) 2020.
170 GA_BALANCE 90.86 % 96.19 % 78.40 % 1 s 1 core @ 2.5 Ghz (Python)
171 MV3D
This method makes use of Velodyne laser scans.
 90.83 % 96.47 % 78.63 % 0.36 s GPU @ 2.5 Ghz (Python + C/C++)
X. Chen, H. Ma, J. Wan, B. Li and T. Xia: Multi-View 3D Object Detection Network for Autonomous Driving. CVPR 2017.
172 3D IoU Loss
This method makes use of Velodyne laser scans.
 90.79 % 95.92 % 85.65 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
D. Zhou, J. Fang, X. Song, C. Guan, J. Yin, Y. Dai and R. Yang: IoU Loss for 2D/3D Object Detection. International Conference on 3D Vision (3DV) 2019.
173 SINet_VGG code 90.79 % 93.59 % 77.53 % 0.2 s TITAN X GPU
X. Hu, X. Xu, Y. Xiao, H. Chen, S. He, J. Qin and P. Heng: SINet: A Scale-insensitive Convolutional Neural Network for Fast Vehicle Detection. IEEE Transactions on Intelligent Transportation Systems 2019.
174 GA_FULLDATA 90.73 % 96.31 % 78.22 % 1 s 4 cores @ 2.5 Ghz (Python)
175 OCM3D 90.70 % 94.36 % 84.56 % 0.5 s 1 core @ 2.5 Ghz (C/C++)
176 Simple3D Net 90.70 % 93.54 % 87.81 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
177 HR-SECOND code 90.68 % 93.72 % 85.63 % 0.11 s 1 core @ 2.5 Ghz (Python + C/C++)
178 GA2500 90.68 % 95.86 % 80.29 % 0.2 s 1 core @ 2.5 Ghz (Python)
179 GA_rpn500 90.68 % 95.86 % 80.29 % 1 s 1 core @ 2.5 Ghz (Python)
180 TANet code 90.67 % 93.67 % 85.31 % 0.035s GPU @ 2.5 Ghz (Python + C/C++)
Z. Liu, X. Zhao, T. Huang, R. Hu, Y. Zhou and X. Bai: TANet: Robust 3D Object Detection from Point Clouds with Triple Attention. AAAI 2020.
181 yolo4 90.63 % 94.71 % 80.38 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
182 baseline 90.59 % 93.29 % 87.18 % 0.12 s 1 core @ 2.5 Ghz (C/C++)
183 VOXEL_FPN_HR 90.55 % 93.76 % 85.42 % 0.12 s 8 cores @ 2.5 Ghz (Python)
ERROR: Wrong syntax in BIBTEX file.
184 Det3D 90.54 % 94.35 % 84.40 % 0.5 s 1 core @ 2.5 Ghz (C/C++)
185 IGRP+ 90.42 % 96.03 % 87.63 % 0.18 s 1 core @ 2.5 Ghz (Python + C/C++)
186 bigger_ga 90.38 % 95.76 % 77.92 % 1 s 1 core @ 2.5 Ghz (Python)
187 CG-Stereo
This method uses stereo information.
 90.38 % 96.31 % 82.80 % 0.57 s GeForce RTX 2080 Ti
C. Li, J. Ku and S. Waslander: Confidence Guided Stereo 3D Object Detection with Split Depth Estimation. IROS 2020.
188 yolo4_5l code 90.38 % 91.79 % 80.64 % 0.02 s 1 core @ 2.5 Ghz (Python + C/C++)
189 SCNet
This method makes use of Velodyne laser scans.
 90.30 % 95.59 % 85.09 % 0.04 s GPU @ 3.0 Ghz (Python)
Z. Wang, H. Fu, L. Wang, L. Xiao and B. Dai: SCNet: Subdivision Coding Network for Object Detection Based on 3D Point Cloud. IEEE Access 2019.
190 RUC code 90.24 % 92.60 % 86.55 % 0.12 s 1 core @ 2.5 Ghz (C/C++)
191 Deep3DBox 90.19 % 94.71 % 76.82 % 1.5 s GPU @ 2.5 Ghz (C/C++)
A. Mousavian, D. Anguelov, J. Flynn and J. Kosecka: 3D Bounding Box Estimation Using Deep Learning and Geometry. CVPR 2017.
192 FQNet 90.17 % 94.72 % 76.78 % 0.5 s 1 core @ 2.5 Ghz (Python)
L. Liu, J. Lu, C. Xu, Q. Tian and J. Zhou: Deep Fitting Degree Scoring Network for Monocular 3D Object Detection. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2019.
193 BVVF 90.15 % 95.65 % 84.95 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
194 DeepStereoOP 90.06 % 95.15 % 79.91 % 3.4 s GPU @ 3.5 Ghz (Matlab + C/C++)
C. Pham and J. Jeon: Robust Object Proposals Re-ranking for Object Detection in Autonomous Driving Using Convolutional Neural Networks. Signal Processing: Image Communiation 2017.
195 SubCNN 89.98 % 94.26 % 79.78 % 2 s GPU @ 3.5 Ghz (Python + C/C++)
Y. Xiang, W. Choi, Y. Lin and S. Savarese: Subcategory-aware Convolutional Neural Networks for Object Proposals and Detection. IEEE Winter Conference on Applications of Computer Vision (WACV) 2017.
196 MLOD
This method makes use of Velodyne laser scans.
 code 89.97 % 94.88 % 84.98 % 0.12 s GPU @ 1.5 Ghz (Python)
J. Deng and K. Czarnecki: MLOD: A multi-view 3D object detection based on robust feature fusion method. arXiv preprint arXiv:1909.04163 2019.
197 GPP code 89.96 % 94.02 % 81.13 % 0.23 s GPU @ 1.5 Ghz (Python + C/C++)
A. Rangesh and M. Trivedi: Ground plane polling for 6dof pose estimation of objects on the road. arXiv preprint arXiv:1811.06666 2018.
198 RUC code 89.93 % 93.12 % 85.44 % 0.12 s 1 core @ 2.5 Ghz (C/C++)
199 AVOD
This method makes use of Velodyne laser scans.
 code 89.88 % 95.17 % 82.83 % 0.08 s Titan X (pascal)
J. Ku, M. Mozifian, J. Lee, A. Harakeh and S. Waslander: Joint 3D Proposal Generation and Object Detection from View Aggregation. IROS 2018.
200 SINet_PVA code 89.86 % 92.72 % 76.47 % 0.11 s TITAN X GPU
X. Hu, X. Xu, Y. Xiao, H. Chen, S. He, J. Qin and P. Heng: SINet: A Scale-insensitive Convolutional Neural Network for Fast Vehicle Detection. IEEE Transactions on Intelligent Transportation Systems 2019.
201 MCA 89.72 % 93.42 % 79.96 % 0.04 s 1 core @ 2.5 Ghz (C/C++)
202 UDI-mono3D 89.67 % 94.39 % 80.29 % 0.05 s 1 core @ 2.5 Ghz (Python)
203 3DOP
This method uses stereo information.
 code 89.55 % 92.96 % 79.38 % 3s GPU @ 2.5 Ghz (Matlab + C/C++)
X. Chen, K. Kundu, Y. Zhu, A. Berneshawi, H. Ma, S. Fidler and R. Urtasun: 3D Object Proposals for Accurate Object Class Detection. NIPS 2015.
204 IAFA 89.46 % 93.08 % 79.83 % 0.04 s 1 core @ 2.5 Ghz (C/C++)
205 Mono3D code 89.37 % 94.52 % 79.15 % 4.2 s GPU @ 2.5 Ghz (Matlab + C/C++)
X. Chen, K. Kundu, Z. Zhang, H. Ma, S. Fidler and R. Urtasun: Monocular 3D Object Detection for Autonomous Driving. CVPR 2016.
206 4d-MSCNN
This method uses stereo information.
 code 89.37 % 92.40 % 77.00 % 0.3 min GPU @ 3.0 Ghz (Matlab + C/C++)
P. Ferraz, B. Oliveira, F. Ferreira, C. Silva Martins and others: Three-stage RGBD architecture for vehicle and pedestrian detection using convolutional neural networks and stereo vision. IET Intelligent Transport Systems 2020.
207 R-FCN(FPN) 89.35 % 93.53 % 79.35 % 0.2 s 1 core @ 2.5 Ghz (Python)
208 Scan_YOLO 88.95 % 90.69 % 79.85 % 0.1 s 4 cores @ 3.0 Ghz (Python)
209 AVOD-FPN
This method makes use of Velodyne laser scans.
 code 88.92 % 94.70 % 84.13 % 0.1 s Titan X (Pascal)
J. Ku, M. Mozifian, J. Lee, A. Harakeh and S. Waslander: Joint 3D Proposal Generation and Object Detection from View Aggregation. IROS 2018.
210 autoRUC 88.88 % 94.23 % 81.35 % 0.12 s 1 core @ 2.5 Ghz (C/C++)
211 Prune 88.85 % 94.20 % 81.31 % 0.11 s 1 core @ 2.5 Ghz (C/C++)
212 AM3D 88.71 % 92.55 % 77.78 % 0.4 s GPU @ 2.5 Ghz (Python + C/C++)
X. Ma, Z. Wang, H. Li, P. Zhang, W. Ouyang and X. Fan: Accurate Monocular Object Detection via Color- Embedded 3D Reconstruction for Autonomous Driving. Proceedings of the IEEE international Conference on Computer Vision (ICCV) 2019.
213 SS3D_HW 88.68 % 94.49 % 68.79 % 0.4 s GPU @ 2.5 Ghz (Python)
214 MS-CNN code 88.68 % 93.87 % 76.11 % 0.4 s GPU @ 2.5 Ghz (C/C++)
Z. Cai, Q. Fan, R. Feris and N. Vasconcelos: A Unified Multi-scale Deep Convolutional Neural Network for Fast Object Detection. ECCV 2016.
215 PMN 88.65 % 93.64 % 77.94 % 0.2 s 1 core @ 2.5 Ghz (Python)
216 CRCNNA 88.59 % 94.82 % 76.74 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
217 MonoPSR code 88.50 % 93.63 % 73.36 % 0.2 s GPU @ 3.5 Ghz (Python)
J. Ku*, A. Pon* and S. Waslander: Monocular 3D Object Detection Leveraging Accurate Proposals and Shape Reconstruction. CVPR 2019.
218 Shift R-CNN (mono) code 88.48 % 94.07 % 78.34 % 0.25 s GPU @ 1.5 Ghz (Python)
A. Naiden, V. Paunescu, G. Kim, B. Jeon and M. Leordeanu: Shift R-CNN: Deep Monocular 3D Object Detection With Closed-form Geometric Constraints. ICIP 2019.
219 PSMD 88.47 % 93.67 % 75.62 % 0.1 s GPU @ 2.5 Ghz (Python)
220 RCD 88.46 % 92.52 % 83.73 % 0.1 s GPU @ 2.5 Ghz (Python)
221 MM-MRFC
This method uses optical flow information.
This method makes use of Velodyne laser scans.
 88.46 % 95.54 % 78.14 % 0.05 s GPU @ 2.5 Ghz (C/C++)
A. Costea, R. Varga and S. Nedevschi: Fast Boosting based Detection using Scale Invariant Multimodal Multiresolution Filtered Features. CVPR 2017.
222 AACL 88.35 % 93.56 % 73.57 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
223 3DBN
This method makes use of Velodyne laser scans.
 88.29 % 93.74 % 80.74 % 0.13s 1080Ti (Python+C/C++)
X. Li, J. Guivant, N. Kwok and Y. Xu: 3D Backbone Network for 3D Object Detection. CoRR 2019.
224 UDI-mono3D 88.16 % 93.93 % 79.57 % 0.05 s 1 core @ 2.5 Ghz (Python)
225 anonymous 88.16 % 96.22 % 75.72 % 1 s 1 core @ 2.5 Ghz (C/C++)
226 CDI3D 87.97 % 91.46 % 80.14 % 0.03 s GPU @ 2.5 Ghz (Python)
227 Multi-task DG 87.72 % 95.50 % 75.51 % 0.06 s GPU @ 2.5 Ghz (Python)
228 ga50 87.65 % 95.76 % 75.14 % 1 s 1 core @ 2.5 Ghz (Python)
229 MMCOM 87.58 % 95.08 % 77.48 % 0.04 s 1 core @ 2.5 Ghz (Python)
230 SMOKE code 87.51 % 93.21 % 77.66 % 0.03 s GPU @ 2.5 Ghz (Python)
Z. Liu, Z. Wu and R. Tóth: SMOKE: Single-Stage Monocular 3D Object Detection via Keypoint Estimation. 2020.
231 tiny-stereo-v2
This method uses stereo information.
 87.46 % 96.16 % 80.12 % 0.4 s 1 core @ 2.5 Ghz (Python + C/C++)
232 DAMNET code 87.39 % 92.48 % 82.41 % 1 s 1 core @ 2.5 Ghz (C/C++)
233 MA 87.29 % 93.21 % 79.82 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
234 CDN
This method uses stereo information.
 87.19 % 95.85 % 79.43 % 0.6 s GPU @ 2.5 Ghz (Python)
D. Garg, Y. Wang, B. Hariharan, M. Campbell, K. Weinberger and W. Chao: Wasserstein Distances for Stereo Disparity Estimation. Advances in Neural Information Processing Systems (NeurIPS) 2020.
235 IMA 87.17 % 92.67 % 77.46 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
236 RTM3D code 86.93 % 91.82 % 77.41 % 0.05 s GPU @ 1.0 Ghz (Python)
P. Li, H. Zhao, P. Liu and F. Cao: RTM3D: Real-time Monocular 3D Detection from Object Keypoints for Autonomous Driving. 2020.
237 yolo_rgb 86.90 % 90.01 % 77.52 % 0.07 s GPU @ 2.5 Ghz (Python)
238 NL_M3D 86.80 % 91.31 % 72.37 % 0.2 s 1 core @ 2.5 Ghz (Python)
239 voxelrcnn 86.69 % 94.60 % 79.91 % 15 s 1 core @ 2.5 Ghz (C/C++)
240 DSGN
This method uses stereo information.
 code 86.43 % 95.53 % 78.75 % 0.67 s NVIDIA Tesla V100
Y. Chen, S. Liu, X. Shen and J. Jia: DSGN: Deep Stereo Geometry Network for 3D Object Detection. CVPR 2020.
241 PB3D
This method uses stereo information.
 86.21 % 95.64 % 76.83 % 0.42 s 1 core @ 2.5 Ghz (C/C++)
242 tiny-stereo-v1
This method uses stereo information.
 86.04 % 95.19 % 78.24 % 0.3 s GPU @ 2.5 Ghz (Python + C/C++)
243 Stereo R-CNN
This method uses stereo information.
 code 85.98 % 93.98 % 71.25 % 0.3 s GPU @ 2.5 Ghz (Python)
P. Li, X. Chen and S. Shen: Stereo R-CNN based 3D Object Detection for Autonomous Driving. CVPR 2019.
244 StereoFENet
This method uses stereo information.
 85.70 % 91.48 % 77.62 % 0.15 s 1 core @ 3.5 Ghz (Python)
W. Bao, B. Xu and Z. Chen: MonoFENet: Monocular 3D Object Detection with Feature Enhancement Networks. IEEE Transactions on Image Processing 2019.
245 ResNet-RRC_Car 85.33 % 91.45 % 74.27 % 0.06 s GPU @ 1.5 Ghz (Python + C/C++)
H. Jeon and others: High-Speed Car Detection Using ResNet- Based Recurrent Rolling Convolution. Proceedings of the IEEE conference on systems, man, and cybernetics 2018.
246 PL++ (SDN+GDC)
This method uses stereo information.
This method makes use of Velodyne laser scans.
 code 85.15 % 94.95 % 77.78 % 0.6 s GPU @ 2.5 Ghz (C/C++)
Y. You, Y. Wang, W. Chao, D. Garg, G. Pleiss, B. Hariharan, M. Campbell and K. Weinberger: Pseudo-LiDAR++: Accurate Depth for 3D Object Detection in Autonomous Driving. International Conference on Learning Representations 2020.
247 RAR-Net 85.08 % 89.04 % 69.26 % 0.5 s 1 core @ 2.5 Ghz (C/C++)
248 M3D-RPN code 85.08 % 89.04 % 69.26 % 0.16 s GPU @ 1.5 Ghz (Python)
G. Brazil and X. Liu: M3D-RPN: Monocular 3D Region Proposal Network for Object Detection . ICCV 2019 .
249 Center3D 85.05 % 95.14 % 73.06 % 0.05 s GPU @ 3.5 Ghz (Python)
250 CDN-PL++
This method uses stereo information.
 85.01 % 94.66 % 77.60 % 0.4 s GPU @ 2.5 Ghz (C/C++)
D. Garg, Y. Wang, B. Hariharan, M. Campbell, K. Weinberger and W. Chao: Wasserstein Distances for Stereo Disparity Estimation. arXiv preprint arXiv:2007.03085 2020.
251 SDP+CRC (ft) 85.00 % 92.06 % 71.71 % 0.6 s GPU @ 2.5 Ghz (C/C++)
F. Yang, W. Choi and Y. Lin: Exploit All the Layers: Fast and Accurate CNN Object Detector with Scale Dependent Pooling and Cascaded Rejection Classifiers. Proceedings of the IEEE International Conference on Computer Vision and Pattern Recognition 2016.
252 bifpn_fsrn 84.93 % 93.68 % 74.45 % 0.07 s 1 core @ 2.5 Ghz (Python + C/C++)
253 ResNet-RRC (pruned) 84.93 % 89.59 % 73.26 % 0.11 s GPU @ 1.5 Ghz (Python + C/C++)
254 IDA-3D
This method uses stereo information.
 84.92 % 92.79 % 74.75 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
255 SS3D 84.92 % 92.72 % 70.35 % 48 ms Tesla V100 (Python)
E. Jörgensen, C. Zach and F. Kahl: Monocular 3D Object Detection and Box Fitting Trained End-to-End Using Intersection-over-Union Loss. CoRR 2019.
256 MP-Mono 84.83 % 91.58 % 65.89 % 0.16 s GPU @ 2.5 Ghz (Python)
257 ResNet-RRC 84.81 % 89.43 % 73.18 % 0.11 s GPU @ 1.5 Ghz (Python + C/C++)
258 MonoFENet 84.63 % 91.68 % 76.71 % 0.15 s 1 core @ 3.5 Ghz (Python)
W. Bao, B. Xu and Z. Chen: MonoFENet: Monocular 3D Object Detection with Feature Enhancement Networks. IEEE Transactions on Image Processing 2019.
259 MV3D (LIDAR)
This method makes use of Velodyne laser scans.
 84.39 % 93.08 % 79.27 % 0.24 s GPU @ 2.5 Ghz (Python + C/C++)
X. Chen, H. Ma, J. Wan, B. Li and T. Xia: Multi-View 3D Object Detection Network for Autonomous Driving. CVPR 2017.
260 Complexer-YOLO
This method makes use of Velodyne laser scans.
 84.16 % 91.92 % 79.62 % 0.06 s GPU @ 3.5 Ghz (C/C++)
M. Simon, K. Amende, A. Kraus, J. Honer, T. Samann, H. Kaulbersch, S. Milz and H. Michael Gross: Complexer-YOLO: Real-Time 3D Object Detection and Tracking on Semantic Point Clouds. The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops 2019.
261 ZoomNet
This method uses stereo information.
 code 83.92 % 94.22 % 69.00 % 0.3 s 1 core @ 2.5 Ghz (C/C++)
L. Z. Xu: ZoomNet: Part-Aware Adaptive Zooming Neural Network for 3D Object Detection. Proceedings of the AAAI Conference on Artificial Intelligence 2020.
262 LAPNet 83.85 % 90.81 % 65.37 % 0.03 s 1 core @ 2.5 Ghz (Python)
263 D4LCN code 83.67 % 90.34 % 65.33 % 0.2 s GPU @ 2.5 Ghz (Python + C/C++)
M. Ding, Y. Huo, H. Yi, Z. Wang, J. Shi, Z. Lu and P. Luo: Learning Depth-Guided Convolutions for Monocular 3D Object Detection. CVPR 2020.
264 seivl 83.60 % 90.35 % 81.76 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
265 ASOD 83.52 % 94.09 % 68.68 % 0.28 s GPU @ 2.5 Ghz (Python)
266 Faster R-CNN code 83.16 % 88.97 % 72.62 % 2 s GPU @ 3.5 Ghz (Python + C/C++)
S. Ren, K. He, R. Girshick and J. Sun: Faster R-CNN: Towards Real- Time Object Detection with Region Proposal Networks. NIPS 2015.
267 Mag 83.15 % 94.24 % 70.63 % 0.07 s 1 core @ 2.5 Ghz (C/C++)
268 MTMono3d 83.11 % 90.55 % 75.48 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
269 SSL-RTM3D Res18 82.97 % 93.35 % 73.11 % 0.02 s GPU @ 2.5 Ghz (Python)
270 vxnet 82.91 % 87.98 % 61.70 % 1 s 1 core @ 2.5 Ghz (C/C++)
271 Pseudo-LiDAR++
This method uses stereo information.
 code 82.90 % 94.46 % 75.45 % 0.4 s GPU @ 2.5 Ghz (Python)
Y. You, Y. Wang, W. Chao, D. Garg, G. Pleiss, B. Hariharan, M. Campbell and K. Weinberger: Pseudo-LiDAR++: Accurate Depth for 3D Object Detection in Autonomous Driving. International Conference on Learning Representations 2020.
272 DP3D 82.81 % 87.85 % 66.80 % 0.05 s GPU @ 1.5 Ghz (Python + C/C++)
273 BS3D 82.72 % 95.35 % 70.01 % 22 ms Titan Xp
N. Gählert, J. Wan, M. Weber, J. Zöllner, U. Franke and J. Denzler: Beyond Bounding Boxes: Using Bounding Shapes for Real-Time 3D Vehicle Detection from Monocular RGB Images. 2019 IEEE Intelligent Vehicles Symposium (IV) 2019.
274 DP3D 82.63 % 87.90 % 66.62 % 0.07 s GPU @ 1.5 Ghz (Python + C/C++)
275 Disp R-CNN
This method uses stereo information.
 code 82.57 % 93.26 % 68.20 % 0.42 s GPU @ 2.5 Ghz (Python + C/C++)
J. Sun, L. Chen, Y. Xie, S. Zhang, Q. Jiang, X. Zhou and H. Bao: Disp R-CNN: Stereo 3D Object Detection via Shape Prior Guided Instance Disparity Estimation. CVPR 2020.
276 Pseudo-LiDAR E2E
This method uses stereo information.
 82.54 % 94.00 % 75.31 % 0.4 s GPU @ 2.5 Ghz (Python)
277 Disp R-CNN (velo)
This method uses stereo information.
 code 82.47 % 93.20 % 68.09 % 0.42 s GPU @ 2.5 Ghz (Python + C/C++)
J. Sun, L. Chen, Y. Xie, S. Zhang, Q. Jiang, X. Zhou and H. Bao: Disp R-CNN: Stereo 3D Object Detection via Shape Prior Guided Instance Disparity Estimation. CVPR 2020.
278 deprecated 82.23 % 92.21 % 67.87 % 1 core @ 2.5 Ghz (C/C++)
279 S3D 82.18 % 91.77 % 67.82 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
280 Stereo3D
This method uses stereo information.
 82.15 % 94.81 % 62.17 % 0.1 s GPU 1080Ti
281 LNET 82.02 % 91.49 % 67.71 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
282 FRCNN+Or code 82.00 % 92.91 % 68.79 % 0.09 s Titan Xp GPU
C. Guindel, D. Martin and J. Armingol: Fast Joint Object Detection and Viewpoint Estimation for Traffic Scene Understanding. IEEE Intelligent Transportation Systems Magazine 2018.
C. Guindel, D. Martin and J. Armingol: Joint Object Detection and Viewpoint Estimation using CNN features. IEEE International Conference on Vehicular Electronics and Safety (ICVES) 2017.
283 CBNet 81.70 % 91.47 % 72.02 % 1 s 4 cores @ 2.5 Ghz (Python)
284 yyyyolo 81.33 % 94.36 % 68.72 % 0.01 s 1 core @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
285 LCD3D 81.25 % 91.29 % 64.55 % 0.03 s GPU @ 2.5 Ghz (Python)
286 A3DODWTDA (image) code 81.25 % 78.96 % 70.56 % 0.8 s GPU @ 3.0 Ghz (Python)
F. Gustafsson and E. Linder-Norén: Automotive 3D Object Detection Without Target Domain Annotations. 2018.
287 RefineNet 81.01 % 91.91 % 65.67 % 0.20 s GPU @ 2.5 Ghz (Matlab + C++)
R. Rajaram, E. Bar and M. Trivedi: RefineNet: Refining Object Detectors for Autonomous Driving. IEEE Transactions on Intelligent Vehicles 2016.
R. Rajaram, E. Bar and M. Trivedi: RefineNet: Iterative Refinement for Accurate Object Localization. Intelligent Transportation Systems Conference 2016.
288 I2BEV 80.73 % 93.61 % 71.09 % 0.63 s GPU @ 2.5 Ghz (Python)
289 UM3D_TUM 80.36 % 92.88 % 65.95 % 0.05 s 1 core @ 2.5 Ghz (Python)
290 3D-GCK 80.19 % 89.55 % 68.08 % 24 ms Tesla V100
N. Gählert, J. Wan, N. Jourdan, J. Finkbeiner, U. Franke and J. Denzler: Single-Shot 3D Detection of Vehicles from Monocular RGB Images via Geometrically Constrained Keypoints in Real-Time. 2020 IEEE Intelligent Vehicles Symposium (IV) 2020.
291 YoloMono3D 79.63 % 92.37 % 59.69 % 0.05 s GPU @ 2.5 Ghz (Python)
292 MMRetina
This method uses stereo information.
This method uses optical flow information.
This method makes use of Velodyne laser scans.
 79.53 % 89.66 % 69.52 % 0.38 s GPU @ 2.5 Ghz (Python)
293 DA-3Ddet 79.47 % 89.49 % 63.04 % 0.4 s GPU @ 2.5 Ghz (Python)
294 ITS-MDPL 79.30 % 92.31 % 71.94 % 0.16 s GPU @ 2.5 Ghz (Python)
295 A3DODWTDA
This method makes use of Velodyne laser scans.
 code 79.15 % 82.98 % 68.30 % 0.08 s GPU @ 3.0 Ghz (Python)
F. Gustafsson and E. Linder-Norén: Automotive 3D Object Detection Without Target Domain Annotations. 2018.
296 MTNAS 78.82 % 88.96 % 67.07 % 0.02 s 1 core @ 2.5 Ghz (python)
297 spLBP 78.66 % 81.66 % 61.69 % 1.5 s 8 cores @ 2.5 Ghz (Matlab + C/C++)
Q. Hu, S. Paisitkriangkrai, C. Shen, A. Hengel and F. Porikli: Fast Detection of Multiple Objects in Traffic Scenes With a Common Detection Framework. IEEE Trans. Intelligent Transportation Systems 2016.
298 3D-SSMFCNN code 78.19 % 77.92 % 69.19 % 0.1 s GPU @ 1.5 Ghz (C/C++)
L. Novak: Vehicle Detection and Pose Estimation for Autonomous Driving. 2017.
299 MonoGRNet code 77.94 % 88.65 % 63.31 % 0.04s NVIDIA P40
Z. Qin, J. Wang and Y. Lu: MonoGRNet: A Geometric Reasoning Network for 3D Object Localization. The Thirty-Third AAAI Conference on Artificial Intelligence (AAAI-19) 2019.
300 Reinspect code 77.48 % 90.27 % 66.73 % 2s 1 core @ 2.5 Ghz (C/C++)
R. Stewart, M. Andriluka and A. Ng: End-to-End People Detection in Crowded Scenes. CVPR 2016.
301 multi-task CNN 77.18 % 86.12 % 68.09 % 25.1 ms GPU @ 2.0 Ghz (Python)
M. Oeljeklaus, F. Hoffmann and T. Bertram: A Fast Multi-Task CNN for Spatial Understanding of Traffic Scenes. IEEE Intelligent Transportation Systems Conference 2018.
302 EACV 77.17 % 89.33 % 67.70 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
303 Regionlets 76.99 % 88.75 % 60.49 % 1 s >8 cores @ 2.5 Ghz (C/C++)
X. Wang, M. Yang, S. Zhu and Y. Lin: Regionlets for Generic Object Detection. T-PAMI 2015.
W. Zou, X. Wang, M. Sun and Y. Lin: Generic Object Detection with Dense Neural Patterns and Regionlets. British Machine Vision Conference 2014.
C. Long, X. Wang, G. Hua, M. Yang and Y. Lin: Accurate Object Detection with Location Relaxation and Regionlets Relocalization. Asian Conference on Computer Vision 2014.
304 3DVP code 76.98 % 84.95 % 65.78 % 40 s 8 cores @ 3.5 Ghz (Matlab + C/C++)
Y. Xiang, W. Choi, Y. Lin and S. Savarese: Data-Driven 3D Voxel Patterns for Object Category Recognition. IEEE Conference on Computer Vision and Pattern Recognition 2015.
305 SubCat code 76.36 % 84.10 % 60.56 % 0.7 s 6 cores @ 3.5 Ghz (Matlab + C/C++)
E. Ohn-Bar and M. Trivedi: Learning to Detect Vehicles by Clustering Appearance Patterns. T-ITS 2015.
306 GS3D 76.35 % 86.23 % 62.67 % 2 s 1 core @ 2.5 Ghz (C/C++)
B. Li, W. Ouyang, L. Sheng, X. Zeng and X. Wang: GS3D: An Efficient 3D Object Detection Framework for Autonomous Driving. IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2019.
307 AOG code 76.24 % 86.08 % 61.51 % 3 s 4 cores @ 2.5 Ghz (Matlab)
T. Wu, B. Li and S. Zhu: Learning And-Or Models to Represent Context and Occlusion for Car Detection and Viewpoint Estimation. TPAMI 2016.
B. Li, T. Wu and S. Zhu: Integrating Context and Occlusion for Car Detection by Hierarchical And-Or Model. ECCV 2014.
308 Pose-RCNN 75.83 % 89.59 % 64.06 % 2 s >8 cores @ 2.5 Ghz (Python)
M. Braun, Q. Rao, Y. Wang and F. Flohr: Pose-RCNN: Joint object detection and pose estimation using 3D object proposals. Intelligent Transportation Systems (ITSC), 2016 IEEE 19th International Conference on 2016.
309 3D FCN
This method makes use of Velodyne laser scans.
 74.65 % 86.74 % 67.85 % >5 s 1 core @ 2.5 Ghz (C/C++)
B. Li: 3D Fully Convolutional Network for Vehicle Detection in Point Cloud. IROS 2017.
310 OC Stereo
This method uses stereo information.
 code 74.60 % 87.39 % 62.56 % 0.35 s 1 core @ 2.5 Ghz (Python + C/C++)
A. Pon, J. Ku, C. Li and S. Waslander: Object-Centric Stereo Matching for 3D Object Detection. ICRA 2020.
311 yolo_depth 74.40 % 88.71 % 65.58 % 0.07 s GPU @ 2.5 Ghz (Python)
312 m-prcnn
This method uses stereo information.
 73.64 % 87.64 % 57.03 % 0.43 s 1 core @ 2.5 Ghz (Python)
313 RTS3D 73.08 % 80.48 % 64.02 % 0.03 s GPU @ 2.5 Ghz (Python)
314 RuiRUC 72.08 % 87.48 % 55.28 % 0.12 s 1 core @ 2.5 Ghz (Python)
315 ANM 71.97 % 87.17 % 55.19 % 0.12 s 1 core @ 2.5 Ghz (Python)
316 Kinematic3D code 71.73 % 89.67 % 54.97 % 0.12 s 1 core @ 1.5 Ghz (C/C++)
G. Brazil, G. Pons-Moll, X. Liu and B. Schiele: Kinematic 3D Object Detection in Monocular Video. ECCV 2020 .
317 AOG-View 71.26 % 85.01 % 55.73 % 3 s 1 core @ 2.5 Ghz (Matlab, C/C++)
B. Li, T. Wu and S. Zhu: Integrating Context and Occlusion for Car Detection by Hierarchical And-Or Model. ECCV 2014.
318 DAM 70.78 % 90.08 % 61.38 % 1 s GPU @ 2.5 Ghz (Python)
319 MV-RGBD-RF
This method makes use of Velodyne laser scans.
 70.70 % 77.89 % 57.41 % 4 s 4 cores @ 2.5 Ghz (C/C++)
A. Gonzalez, D. Vazquez, A. Lopez and J. Amores: On-Board Object Detection: Multicue, Multimodal, and Multiview Random Forest of Local Experts.. IEEE Trans. on Cybernetics 2016.
A. Gonzalez, G. Villalonga, J. Xu, D. Vazquez, J. Amores and A. Lopez: Multiview Random Forest of Local Experts Combining RGB and LIDAR data for Pedestrian Detection. IEEE Intelligent Vehicles Symposium (IV) 2015.
320 Vote3Deep
This method makes use of Velodyne laser scans.
 70.30 % 78.95 % 63.12 % 1.5 s 4 cores @ 2.5 Ghz (C/C++)
M. Engelcke, D. Rao, D. Zeng Wang, C. Hay Tong and I. Posner: Vote3Deep: Fast Object Detection in 3D Point Clouds Using Efficient Convolutional Neural Networks. ArXiv e-prints 2016.
321 ROI-10D 70.16 % 76.56 % 61.15 % 0.2 s GPU @ 3.5 Ghz (Python)
F. Manhardt, W. Kehl and A. Gaidon: ROI-10D: Monocular Lifting of 2D Detection to 6D Pose and Metric Shape. Computer Vision and Pattern Recognition (CVPR) 2019.
322 BirdNet+
This method makes use of Velodyne laser scans.
 code 68.05 % 92.10 % 65.61 % 0.1 s Titan Xp (PyTorch)
A. Barrera, C. Guindel, J. Beltrán and F. García: BirdNet+: End-to-End 3D Object Detection in LiDAR Bird's Eye View. arXiv:2003.04188 [cs.CV] 2020.
323 Decoupled-3D 67.92 % 87.78 % 54.53 % 0.08 s GPU @ 2.5 Ghz (C/C++)
Y. Cai, B. Li, Z. Jiao, H. Li, X. Zeng and X. Wang: Monocular 3D Object Detection with Decoupled Structured Polygon Estimation and Height-Guided Depth Estimation. AAAI 2020.
324 Pseudo-Lidar
This method uses stereo information.
 code 67.79 % 85.40 % 58.50 % 0.4 s GPU @ 2.5 Ghz (Python + C/C++)
Y. Wang, W. Chao, D. Garg, B. Hariharan, M. Campbell and K. Weinberger: Pseudo-LiDAR From Visual Depth Estimation: Bridging the Gap in 3D Object Detection for Autonomous Driving. The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2019.
325 OC-DPM 67.06 % 79.07 % 52.61 % 10 s 8 cores @ 2.5 Ghz (Matlab)
B. Pepik, M. Stark, P. Gehler and B. Schiele: Occlusion Patterns for Object Class Detection. IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2013.
326 DPM-VOC+VP 66.72 % 82.15 % 49.01 % 8 s 1 core @ 2.5 Ghz (C/C++)
B. Pepik, M. Stark, P. Gehler and B. Schiele: Multi-view and 3D Deformable Part Models. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) 2015.
327 BdCost48LDCF code 66.63 % 81.38 % 52.20 % 0.5 s 8 cores @ 3.5 Ghz (Matlab + C/C++)
A. Fernández-Baldera, J. Buenaposada and L. Baumela: BAdaCost: Multi-class Boosting with Costs . Pattern Recognition 2018.
328 RefinedMPL 65.24 % 88.29 % 53.20 % 0.15 s GPU @ 2.5 Ghz (Python + C/C++)
J. Vianney, S. Aich and B. Liu: RefinedMPL: Refined Monocular PseudoLiDAR for 3D Object Detection in Autonomous Driving. arXiv preprint arXiv:1911.09712 2019.
329 MDPM-un-BB 64.06 % 79.74 % 49.07 % 60 s 4 core @ 2.5 Ghz (MATLAB)
P. Felzenszwalb, R. Girshick, D. McAllester and D. Ramanan: Object Detection with Discriminatively Trained Part-Based Models. PAMI 2010.
330 TLNet (Stereo)
This method uses stereo information.
 code 63.53 % 76.92 % 54.58 % 0.1 s 1 core @ 2.5 Ghz (Python)
Z. Qin, J. Wang and Y. Lu: Triangulation Learning Network: from Monocular to Stereo 3D Object Detection. IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2019.
331 PDV-Subcat 63.24 % 78.27 % 47.67 % 7 s 1 core @ 2.5 Ghz (C/C++)
J. Shen, X. Zuo, J. Li, W. Yang and H. Ling: A novel pixel neighborhood differential statistic feature for pedestrian and face detection . Pattern Recognition 2017.
332 PG-MonoNet 62.75 % 70.87 % 54.34 % 0.19 s GPU @ 2.5 Ghz (Python)
333 MODet
This method makes use of Velodyne laser scans.
 62.54 % 66.06 % 60.04 % 0.05 s GTX1080Ti
Y. Zhang, Z. Xiang, C. Qiao and S. Chen: Accurate and Real-Time Object Detection Based on Bird's Eye View on 3D Point Clouds. 2019 International Conference on 3D Vision (3DV) 2019.
334 SubCat48LDCF code 61.16 % 78.86 % 44.69 % 0.5 s 8 cores @ 3.5 Ghz (Matlab + C/C++)
A. Fernández-Baldera, J. Buenaposada and L. Baumela: BAdaCost: Multi-class Boosting with Costs . Pattern Recognition 2018.
335 DPM-C8B1
This method uses stereo information.
 60.21 % 75.24 % 44.73 % 15 s 4 cores @ 2.5 Ghz (Matlab + C/C++)
J. Yebes, L. Bergasa and M. García-Garrido: Visual Object Recognition with 3D-Aware Features in KITTI Urban Scenes. Sensors 2015.
J. Yebes, L. Bergasa, R. Arroyo and A. Lázaro: Supervised learning and evaluation of KITTI's cars detector with DPM. IV 2014.
336 SAMME48LDCF code 58.38 % 77.47 % 44.43 % 0.5 s 8 cores @ 3.5 Ghz (Matlab + C/C++)
A. Fernández-Baldera, J. Buenaposada and L. Baumela: BAdaCost: Multi-class Boosting with Costs . Pattern Recognition 2018.
337 LSVM-MDPM-sv 58.36 % 71.11 % 43.22 % 10 s 4 cores @ 3.0 Ghz (C/C++)
P. Felzenszwalb, R. Girshick, D. McAllester and D. Ramanan: Object Detection with Discriminatively Trained Part-Based Models. PAMI 2010.
A. Geiger, C. Wojek and R. Urtasun: Joint 3D Estimation of Objects and Scene Layout. NIPS 2011.
338 BirdNet
This method makes use of Velodyne laser scans.
 57.12 % 79.30 % 55.16 % 0.11 s Titan Xp (Caffe)
J. Beltrán, C. Guindel, F. Moreno, D. Cruzado, F. García and A. Escalera: BirdNet: A 3D Object Detection Framework from LiDAR Information. 2018 21st International Conference on Intelligent Transportation Systems (ITSC) 2018.
339 ACF-SC 56.60 % 69.90 % 43.61 % <0.3 s 1 core @ >3.5 Ghz (Matlab + C/C++)
C. Cadena, A. Dick and I. Reid: A Fast, Modular Scene Understanding System using Context-Aware Object Detection. Robotics and Automation (ICRA), 2015 IEEE International Conference on 2015.
340 LSVM-MDPM-us code 55.95 % 68.94 % 41.45 % 10 s 4 cores @ 3.0 Ghz (C/C++)
P. Felzenszwalb, R. Girshick, D. McAllester and D. Ramanan: Object Detection with Discriminatively Trained Part-Based Models. PAMI 2010.
341 ACF 54.09 % 63.05 % 41.81 % 0.2 s 1 core @ >3.5 Ghz (Matlab + C/C++)
P. Doll\'ar, R. Appel, S. Belongie and P. Perona: Fast Feature Pyramids for Object Detection. PAMI 2014.
P. Doll\'ar: Piotr's Image and Video Matlab Toolbox (PMT). .
342 Mono3D_PLiDAR code 53.36 % 80.85 % 44.80 % 0.1 s NVIDIA GeForce 1080 (pytorch)
X. Weng and K. Kitani: Monocular 3D Object Detection with Pseudo-LiDAR Point Cloud. arXiv:1903.09847 2019.
343 RT3D-GMP
This method uses stereo information.
 51.95 % 62.41 % 39.14 % 0.06 s GPU @ 2.5 Ghz (Python + C/C++)
344 VeloFCN
This method makes use of Velodyne laser scans.
 51.82 % 70.53 % 45.70 % 1 s GPU @ 2.5 Ghz (Python + C/C++)
B. Li, T. Zhang and T. Xia: Vehicle Detection from 3D Lidar Using Fully Convolutional Network. RSS 2016 .
345 SparVox3D 50.28 % 53.51 % 40.38 % 0.05 s GPU @ 2.0 Ghz (Python)
346 SF
This method uses stereo information.
This method makes use of Velodyne laser scans.
 46.68 % 60.62 % 38.22 % 0.5 s 1 core @ 2.5 Ghz (Python)
ERROR: Wrong syntax in BIBTEX file.
347 Vote3D
This method makes use of Velodyne laser scans.
 45.94 % 54.38 % 40.48 % 0.5 s 4 cores @ 2.8 Ghz (C/C++)
D. Wang and I. Posner: Voting for Voting in Online Point Cloud Object Detection. Proceedings of Robotics: Science and Systems 2015.
348 TopNet-HighRes
This method makes use of Velodyne laser scans.
 45.85 % 58.04 % 41.11 % 101ms NVIDIA GeForce 1080 Ti (tensorflow-gpu)
S. Wirges, T. Fischer, C. Stiller and J. Frias: Object Detection and Classification in Occupancy Grid Maps Using Deep Convolutional Networks. 2018 21st International Conference on Intelligent Transportation Systems (ITSC) 2018.
349 RT3DStereo
This method uses stereo information.
 45.81 % 56.53 % 37.63 % 0.08 s GPU @ 2.5 Ghz (C/C++)
H. Königshof, N. Salscheider and C. Stiller: Realtime 3D Object Detection for Automated Driving Using Stereo Vision and Semantic Information. Proc. IEEE Intl. Conf. Intelligent Transportation Systems 2019.
350 Multimodal Detection
This method makes use of Velodyne laser scans.
 code 45.46 % 63.91 % 37.25 % 0.06 s GPU @ 3.5 Ghz (Matlab + C/C++)
A. Asvadi, L. Garrote, C. Premebida, P. Peixoto and U. Nunes: Multimodal vehicle detection: fusing 3D- LIDAR and color camera data. Pattern Recognition Letters 2017.
351 RT3D
This method makes use of Velodyne laser scans.
 39.69 % 50.33 % 40.04 % 0.09 s GPU @ 1.8Ghz
Y. Zeng, Y. Hu, S. Liu, J. Ye, Y. Han, X. Li and N. Sun: RT3D: Real-Time 3-D Vehicle Detection in LiDAR Point Cloud for Autonomous Driving. IEEE Robotics and Automation Letters 2018.
352 VoxelJones code 36.31 % 43.89 % 34.16 % .18 s 1 core @ 2.5 Ghz (Python + C/C++)
M. Motro and J. Ghosh: Vehicular Multi-object Tracking with Persistent Detector Failures. arXiv preprint arXiv:1907.11306 2019.
353 CSoR
This method makes use of Velodyne laser scans.
 code 21.66 % 31.52 % 17.99 % 3.5 s 4 cores @ >3.5 Ghz (Python + C/C++)
L. Plotkin: PyDriver: Entwicklung eines Frameworks für räumliche Detektion und Klassifikation von Objekten in Fahrzeugumgebung. 2015.
354 mBoW
This method makes use of Velodyne laser scans.
 21.59 % 35.22 % 16.89 % 10 s 1 core @ 2.5 Ghz (C/C++)
J. Behley, V. Steinhage and A. Cremers: Laser-based Segment Classification Using a Mixture of Bag-of-Words. Proc. of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2013.
355 DepthCN
This method makes use of Velodyne laser scans.
 code 21.18 % 37.45 % 16.08 % 2.3 s GPU @ 3.5 Ghz (Matlab)
A. Asvadi, L. Garrote, C. Premebida, P. Peixoto and U. Nunes: DepthCN: vehicle detection using 3D- LIDAR and convnet. IEEE ITSC 2017.
356 YOLOv2 code 14.31 % 26.74 % 10.94 % 0.02 s GPU @ 3.5 Ghz (C/C++)
J. Redmon, S. Divvala, R. Girshick and A. Farhadi: You only look once: Unified, real-time object detection. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2016.
J. Redmon and A. Farhadi: YOLO9000: Better, Faster, Stronger. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2017.
357 TopNet-UncEst
This method makes use of Velodyne laser scans.
 6.24 % 7.24 % 5.42 % 0.09 s NVIDIA GeForce 1080 Ti (tensorflow-gpu)
S. Wirges, M. Braun, M. Lauer and C. Stiller: Capturing Object Detection Uncertainty in Multi-Layer Grid Maps. 2019.
358 TopNet-Retina
This method makes use of Velodyne laser scans.
 5.00 % 6.82 % 4.52 % 52ms GeForce 1080Ti (tensorflow-gpu, v1.12)
S. Wirges, T. Fischer, C. Stiller and J. Frias: Object Detection and Classification in Occupancy Grid Maps Using Deep Convolutional Networks. 2018 21st International Conference on Intelligent Transportation Systems (ITSC) 2018.
359 ANM 0.01 % 0.01 % 0.02 % 0.12 s 1 core @ 2.5 Ghz (C/C++)
360 TopNet-DecayRate
This method makes use of Velodyne laser scans.
 0.01 % 0.00 % 0.01 % 92 ms NVIDIA GeForce 1080 Ti (tensorflow-gpu)
S. Wirges, T. Fischer, C. Stiller and J. Frias: Object Detection and Classification in Occupancy Grid Maps Using Deep Convolutional Networks. 2018 21st International Conference on Intelligent Transportation Systems (ITSC) 2018.
361 LaserNet 0.00 % 0.00 % 0.00 % 12 ms GPU @ 2.5 Ghz (C/C++)
G. Meyer, A. Laddha, E. Kee, C. Vallespi-Gonzalez and C. Wellington: LaserNet: An Efficient Probabilistic 3D Object Detector for Autonomous Driving. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2019.
Table as LaTeX | Only published Methods

Pedestrian


Method Setting Code Moderate Easy Hard Runtime Environment
1 HWFD 83.06 % 90.50 % 78.35 % 0.21 s one 1080Ti
2 dgist_multiDetNet 80.21 % 89.21 % 75.77 % 0.08 s GPU Titanx Pascal (Python)
3 F-PointNet
This method makes use of Velodyne laser scans.
 code 80.13 % 89.83 % 75.05 % 0.17 s GPU @ 3.0 Ghz (Python)
C. Qi, W. Liu, C. Wu, H. Su and L. Guibas: Frustum PointNets for 3D Object Detection from RGB-D Data. arXiv preprint arXiv:1711.08488 2017.
4 TuSimple code 78.40 % 88.87 % 73.66 % 1.6 s GPU @ 2.5 Ghz (Python + C/C++)
F. Yang, W. Choi and Y. Lin: Exploit all the layers: Fast and accurate cnn object detector with scale dependent pooling and cascaded rejection classifiers. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2016.
K. He, X. Zhang, S. Ren and J. Sun: Deep residual learning for image recognition. Proceedings of the IEEE conference on computer vision and pattern recognition 2016.
5 RRC code 76.61 % 85.98 % 71.47 % 3.6 s GPU @ 2.5 Ghz (C/C++)
J. Ren, X. Chen, J. Liu, W. Sun, J. Pang, Q. Yan, Y. Tai and L. Xu: Accurate Single Stage Detector Using Recurrent Rolling Convolution. CVPR 2017.
6 WSSN
This method makes use of Velodyne laser scans.
 76.42 % 84.91 % 71.86 % 0.41 s GPU @ >3.5 Ghz (Python + C/C++)
7 ECP Faster R-CNN 76.25 % 85.96 % 70.55 % 0.25 s GPU @ 2.5 Ghz (Python)
M. Braun, S. Krebs, F. Flohr and D. Gavrila: The EuroCity Persons Dataset: A Novel Benchmark for Object Detection. CoRR 2018.
8 Aston-EAS 76.07 % 86.71 % 70.02 % 0.24 s GPU @ 2.5 Ghz (Python + C/C++)
J. Wei, J. He, Y. Zhou, K. Chen, Z. Tang and Z. Xiong: Enhanced Object Detection With Deep Convolutional Neural Networks for Advanced Driving Assistance. IEEE Transactions on Intelligent Transportation Systems 2019.
9 MHN 75.99 % 87.21 % 69.50 % 0.39 s GPU @ 2.5 Ghz (Python)
J. Cao, Y. Pang, S. Zhao and X. Li: High-Level Semantic Networks for Multi- Scale Object Detection. IEEE Transactions on Circuits and Systems for Video Technology 2019.
10 FFNet code 75.81 % 87.17 % 69.86 % 1.07 s GPU @ 1.5 Ghz (Python)
C. Zhao, Y. Qian and M. Yang: Monocular Pedestrian Orientation Estimation Based on Deep 2D-3D Feedforward. Pattern Recognition 2019.
11 SJTU-HW 75.81 % 87.17 % 69.86 % 0.85s GPU @ 1.5 Ghz (Python + C/C++)
S. Zhang, X. Zhao, L. Fang, F. Haiping and S. Haitao: LED: LOCALIZATION-QUALITY ESTIMATION EMBEDDED DETECTOR. IEEE International Conference on Image Processing 2018.
L. Fang, X. Zhao and S. Zhang: Small-objectness sensitive detection based on shifted single shot detector. Multimedia Tools and Applications 2018.
12 Noah CV Lab - SSL 75.64 % 86.57 % 70.53 % 0.1 s GPU @ 2.5 Ghz (Python)
13 Faster RCNN + Gr + A 74.95 % 86.95 % 69.50 % 1.29 s GPU @ 2.5 Ghz (Python)
14 MS-CNN code 74.89 % 85.71 % 68.99 % 0.4 s GPU @ 2.5 Ghz (C/C++)
Z. Cai, Q. Fan, R. Feris and N. Vasconcelos: A Unified Multi-scale Deep Convolutional Neural Network for Fast Object Detection. ECCV 2016.
15 Faster RCNN + G 73.75 % 85.51 % 68.54 % 1.1 s GPU @ 2.5 Ghz (Python)
16 MMCOM 73.08 % 86.01 % 68.38 % 0.04 s 1 core @ 2.5 Ghz (Python)
17 F-ConvNet
This method makes use of Velodyne laser scans.
 code 72.91 % 83.63 % 67.18 % 0.47 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Wang and K. Jia: Frustum ConvNet: Sliding Frustums to Aggregate Local Point-Wise Features for Amodal 3D Object Detection. IROS 2019.
18 Faster RCNN + A 72.67 % 86.21 % 67.55 % 0.19 s GPU @ 2.5 Ghz (Python)
19 GN 72.29 % 82.93 % 65.56 % 1 s GPU @ 2.5 Ghz (Matlab + C/C++)
S. Jung and K. Hong: Deep network aided by guiding network for pedestrian detection. Pattern Recognition Letters 2017.
20 SubCNN 72.27 % 84.88 % 66.82 % 2 s GPU @ 3.5 Ghz (Python + C/C++)
Y. Xiang, W. Choi, Y. Lin and S. Savarese: Subcategory-aware Convolutional Neural Networks for Object Proposals and Detection. IEEE Winter Conference on Applications of Computer Vision (WACV) 2017.
21 FRCNN-WS 72.26 % 84.20 % 67.47 % 0.22 s 1 core @ 3.0 Ghz (Python)
22 HR-faster-rcnn 72.26 % 87.65 % 65.71 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
23 Faster RCNN + A 72.09 % 85.35 % 66.87 % 0.19 s GPU @ 2.5 Ghz (Python + C/C++)
24 VMVS
This method makes use of Velodyne laser scans.
 71.82 % 82.80 % 66.85 % 0.25 s GPU @ 2.5 Ghz (Python)
J. Ku, A. Pon, S. Walsh and S. Waslander: Improving 3D object detection for pedestrians with virtual multi-view synthesis orientation estimation. IROS 2019.
25 Multi-task DG 71.64 % 85.34 % 66.76 % 0.06 s GPU @ 2.5 Ghz (Python)
26 IVA code 71.37 % 84.61 % 64.90 % 0.4 s GPU @ 2.5 Ghz (C/C++)
Y. Zhu, J. Wang, C. Zhao, H. Guo and H. Lu: Scale-adaptive Deconvolutional Regression Network for Pedestrian Detection. ACCV 2016.
S. Ren, K. He, R. Girshick and J. Sun: Faster R-CNN: Towards real-time object detection with region proposal networks. Advances in neural information processing systems 2015.
27 MM-MRFC
This method uses optical flow information.
This method makes use of Velodyne laser scans.
 70.76 % 83.79 % 64.81 % 0.05 s GPU @ 2.5 Ghz (C/C++)
A. Costea, R. Varga and S. Nedevschi: Fast Boosting based Detection using Scale Invariant Multimodal Multiresolution Filtered Features. CVPR 2017.
28 SDP+RPN 70.42 % 82.07 % 65.09 % 0.4 s GPU @ 2.5 Ghz (Python + C/C++)
F. Yang, W. Choi and Y. Lin: Exploit All the Layers: Fast and Accurate CNN Object Detector with Scale Dependent Pooling and Cascaded Rejection Classifiers. Proceedings of the IEEE International Conference on Computer Vision and Pattern Recognition 2016.
S. Ren, K. He, R. Girshick and J. Sun: Faster R-CNN: Towards real-time object detection with region proposal networks. Advances in Neural Information Processing Systems 2015.
29 Mono3CN 69.75 % 83.47 % 63.15 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
30 3DOP
This method uses stereo information.
 code 69.57 % 83.17 % 63.48 % 3s GPU @ 2.5 Ghz (Matlab + C/C++)
X. Chen, K. Kundu, Y. Zhu, A. Berneshawi, H. Ma, S. Fidler and R. Urtasun: 3D Object Proposals for Accurate Object Class Detection. NIPS 2015.
31 MonoPSR code 68.56 % 85.60 % 63.34 % 0.2 s GPU @ 3.5 Ghz (Python)
J. Ku*, A. Pon* and S. Waslander: Monocular 3D Object Detection Leveraging Accurate Proposals and Shape Reconstruction. CVPR 2019.
32 DeepStereoOP 68.46 % 83.00 % 63.35 % 3.4 s GPU @ 3.5 Ghz (Matlab + C/C++)
C. Pham and J. Jeon: Robust Object Proposals Re-ranking for Object Detection in Autonomous Driving Using Convolutional Neural Networks. Signal Processing: Image Communiation 2017.
33 sensekitti code 68.41 % 82.72 % 62.72 % 4.5 s GPU @ 2.5 Ghz (Python + C/C++)
B. Yang, J. Yan, Z. Lei and S. Li: Craft Objects from Images. CVPR 2016.
34 FII-CenterNet 67.31 % 81.32 % 61.29 % 0.09 s GPU @ 2.5 Ghz (Python)
35 Mono3D code 67.29 % 80.30 % 62.23 % 4.2 s GPU @ 2.5 Ghz (Matlab + C/C++)
X. Chen, K. Kundu, Z. Zhang, H. Ma, S. Fidler and R. Urtasun: Monocular 3D Object Detection for Autonomous Driving. CVPR 2016.
36 Faster R-CNN code 66.24 % 79.97 % 61.09 % 2 s GPU @ 3.5 Ghz (Python + C/C++)
S. Ren, K. He, R. Girshick and J. Sun: Faster R-CNN: Towards Real- Time Object Detection with Region Proposal Networks. NIPS 2015.
37 PMN 66.17 % 82.16 % 60.84 % 0.2 s 1 core @ 2.5 Ghz (Python)
38 SDP+CRC (ft) 64.36 % 79.22 % 59.16 % 0.6 s GPU @ 2.5 Ghz (C/C++)
F. Yang, W. Choi and Y. Lin: Exploit All the Layers: Fast and Accurate CNN Object Detector with Scale Dependent Pooling and Cascaded Rejection Classifiers. Proceedings of the IEEE International Conference on Computer Vision and Pattern Recognition 2016.
39 CRCNNA 63.69 % 78.10 % 58.41 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
40 Pose-RCNN 63.54 % 80.07 % 57.02 % 2 s >8 cores @ 2.5 Ghz (Python)
M. Braun, Q. Rao, Y. Wang and F. Flohr: Pose-RCNN: Joint object detection and pose estimation using 3D object proposals. Intelligent Transportation Systems (ITSC), 2016 IEEE 19th International Conference on 2016.
41 UDI-mono3D 63.24 % 77.94 % 57.35 % 0.05 s 1 core @ 2.5 Ghz (Python)
42 CFM 62.84 % 74.76 % 56.06 % <2 s GPU @ 2.5 Ghz (Matlab + C/C++)
Q. Hu, P. Wang, C. Shen, A. Hengel and F. Porikli: Pushing the Limits of Deep CNNs for Pedestrian Detection. IEEE Transactions on Circuits and Systems for Video Technology 2017.
43 HotSpotNet 62.31 % 71.43 % 59.24 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
Q. Chen, L. Sun, Z. Wang, K. Jia and A. Yuille: object as hotspots. Proceedings of the European Conference on Computer Vision (ECCV) 2020.
44 UDI-mono3D 62.26 % 77.16 % 56.39 % 0.05 s 1 core @ 2.5 Ghz (Python)
45 MonoPair 61.57 % 78.81 % 56.51 % 0.06 s GPU @ 2.5 Ghz (Python + C/C++)
Y. Chen, L. Tai, K. Sun and M. Li: MonoPair: Monocular 3D Object Detection Using Pairwise Spatial Relationships. The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2020.
46 RPN+BF code 61.22 % 77.06 % 55.22 % 0.6 s GPU @ 2.5 Ghz (Matlab + C/C++)
L. Zhang, L. Lin, X. Liang and K. He: Is Faster R-CNN Doing Well for Pedestrian Detection?. ECCV 2016.
47 JSU-NET 61.19 % 83.17 % 56.20 % 0.1 s 1 core @ 2.5 Ghz (Python)
48 RethinkDet3D 60.88 % 70.56 % 56.69 % 0.15 s 1 core @ 2.5 Ghz (Python)
49 Regionlets 60.83 % 73.79 % 54.72 % 1 s >8 cores @ 2.5 Ghz (C/C++)
X. Wang, M. Yang, S. Zhu and Y. Lin: Regionlets for Generic Object Detection. T-PAMI 2015.
W. Zou, X. Wang, M. Sun and Y. Lin: Generic Object Detection with Dense Neural Patterns and Regionlets. British Machine Vision Conference 2014.
C. Long, X. Wang, G. Hua, M. Yang and Y. Lin: Accurate Object Detection with Location Relaxation and Regionlets Relocalization. Asian Conference on Computer Vision 2014.
50 3DSSD code 60.51 % 72.33 % 56.28 % 0.04 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Yang, Y. Sun, S. Liu and J. Jia: 3DSSD: Point-based 3D Single Stage Object Detector. CVPR 2020.
51 MVX-Net++ 60.21 % 69.70 % 56.07 % 0.15 s 1 core @ 2.5 Ghz (C/C++)
52 VICNet 60.07 % 70.41 % 56.29 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
53 PiP 59.94 % 70.52 % 56.51 % 0.033 s 1 core @ 2.5 Ghz (Python)
54 TANet code 59.07 % 69.90 % 56.44 % 0.035s GPU @ 2.5 Ghz (Python + C/C++)
Z. Liu, X. Zhao, T. Huang, R. Hu, Y. Zhou and X. Bai: TANet: Robust 3D Object Detection from Point Clouds with Triple Attention. AAAI 2020.
55 CA-RCNN 58.89 % 68.54 % 56.08 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
56 Deformable PV-RCNN
This method makes use of Velodyne laser scans.
 code 58.81 % 66.93 % 56.57 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
P. Bhattacharyya and K. Czarnecki: Deformable PV-RCNN: Improving 3D Object Detection with Learned Deformations. ECCV 2020 Perception for Autonomous Driving Workshop.
57 SVGA-Net
This method makes use of Velodyne laser scans.
 58.70 % 68.45 % 56.23 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
58 Baseline of CA RCNN 58.68 % 68.44 % 56.22 % 0.1 s GPU @ 2.5 Ghz (Python)
59 CVIS-DF3D 58.68 % 68.44 % 56.22 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
60 AP-RCNN 58.65 % 70.40 % 55.02 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
61 MMLab PV-RCNN
This method makes use of Velodyne laser scans.
 code 58.37 % 68.88 % 55.38 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
S. Shi, C. Guo, L. Jiang, Z. Wang, J. Shi, X. Wang and H. Li: PV-RCNN: Point-Voxel Feature Set Abstraction for 3D Object Detection. CVPR 2020.
62 Point-GNN
This method makes use of Velodyne laser scans.
 code 58.20 % 71.59 % 54.06 % 0.6 s GPU @ 2.5 Ghz (Python)
W. Shi and R. Rajkumar: Point-GNN: Graph Neural Network for 3D Object Detection in a Point Cloud. CVPR 2020.
63 PP-3D 58.20 % 71.59 % 54.06 % 0.1 s 1 core @ 2.5 Ghz (Python)
64 DeepParts 58.15 % 71.47 % 51.92 % ~1 s GPU @ 2.5 Ghz (Matlab)
Y. Tian, P. Luo, X. Wang and X. Tang: Deep Learning Strong Parts for Pedestrian Detection. ICCV 2015.
65 CompACT-Deep 58.14 % 70.93 % 52.29 % 1 s 1 core @ 2.5 Ghz (Matlab + C/C++)
Z. Cai, M. Saberian and N. Vasconcelos: Learning Complexity-Aware Cascades for Deep Pedestrian Detection. ICCV 2015.
66 PPBA 58.06 % 67.73 % 55.69 % NA s GPU @ 2.5 Ghz (Python)
67 MMLab-PartA^2
This method makes use of Velodyne laser scans.
 code 57.96 % 68.78 % 54.01 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
S. Shi, Z. Wang, J. Shi, X. Wang and H. Li: From Points to Parts: 3D Object Detection from Point Cloud with Part-aware and Part-aggregation Network. IEEE Transactions on Pattern Analysis and Machine Intelligence 2020.
68 AVOD-FPN
This method makes use of Velodyne laser scans.
 code 57.87 % 67.95 % 55.23 % 0.1 s Titan X (Pascal)
J. Ku, M. Mozifian, J. Lee, A. Harakeh and S. Waslander: Joint 3D Proposal Generation and Object Detection from View Aggregation. IROS 2018.
69 TBU 57.44 % 67.29 % 54.00 % NA s GPU @ 2.5 Ghz (Python)
70 CentrNet-FG 57.40 % 68.27 % 54.11 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
71 SemanticVoxels 57.22 % 67.62 % 54.90 % 0.04 s GPU @ 2.5 Ghz (Python + C/C++)
J. Fei, W. Chen, P. Heidenreich, S. Wirges and C. Stiller: SemanticVoxels: Sequential Fusion for 3D Pedestrian Detection using LiDAR Point Cloud and Semantic Segmentation. MFI 2020.
72 IC-PVRCNN 57.20 % 65.43 % 55.17 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
73 IC-SECOND 57.20 % 65.53 % 54.92 % 0.06 s 1 core @ 2.5 Ghz (C/C++)
74 Simple3D Net 57.00 % 66.89 % 54.38 % 0.02 s GPU @ 2.5 Ghz (Python)
75 KNN-GCNN 56.80 % 69.53 % 52.86 % 0.4 s 1 core @ 2.5 Ghz (Python + C/C++)
76 FRCNN+Or code 56.68 % 71.64 % 51.53 % 0.09 s Titan Xp GPU
C. Guindel, D. Martin and J. Armingol: Fast Joint Object Detection and Viewpoint Estimation for Traffic Scene Understanding. IEEE Intelligent Transportation Systems Magazine 2018.
C. Guindel, D. Martin and J. Armingol: Joint Object Detection and Viewpoint Estimation using CNN features. IEEE International Conference on Vehicular Electronics and Safety (ICVES) 2017.
77 FilteredICF 56.53 % 69.79 % 50.32 % ~ 2 s >8 cores @ 2.5 Ghz (Matlab + C/C++)
S. Zhang, R. Benenson and B. Schiele: Filtered Channel Features for Pedestrian Detection. CVPR 2015.
78 yolo4_5l 56.46 % 73.14 % 49.57 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
79 ARPNET 56.42 % 69.08 % 52.69 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
Y. Ye, C. Zhang and X. Hao: ARPNET: attention region proposal network for 3D object detection. Science China Information Sciences 2019.
80 MV-RGBD-RF
This method makes use of Velodyne laser scans.
 56.18 % 72.99 % 49.72 % 4 s 4 cores @ 2.5 Ghz (C/C++)
A. Gonzalez, D. Vazquez, A. Lopez and J. Amores: On-Board Object Detection: Multicue, Multimodal, and Multiview Random Forest of Local Experts.. IEEE Trans. on Cybernetics 2016.
A. Gonzalez, G. Villalonga, J. Xu, D. Vazquez, J. Amores and A. Lopez: Multiview Random Forest of Local Experts Combining RGB and LIDAR data for Pedestrian Detection. IEEE Intelligent Vehicles Symposium (IV) 2015.
81 CVIS-DF3D_v2 56.02 % 65.82 % 53.58 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
82 yolo4 55.78 % 72.49 % 51.11 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
83 Mag 55.74 % 71.91 % 50.92 % 0.07 s 1 core @ 2.5 Ghz (C/C++)
84 MLOD
This method makes use of Velodyne laser scans.
 code 55.62 % 68.42 % 51.45 % 0.12 s GPU @ 1.5 Ghz (Python)
J. Deng and K. Czarnecki: MLOD: A multi-view 3D object detection based on robust feature fusion method. arXiv preprint arXiv:1909.04163 2019.
85 DAM 55.60 % 74.85 % 50.63 % 1 s GPU @ 2.5 Ghz (Python)
86 CDI3D 55.16 % 67.35 % 51.17 % 0.03 s GPU @ 2.5 Ghz (Python)
87 PointPillars
This method makes use of Velodyne laser scans.
 code 55.10 % 65.29 % 52.39 % 16 ms 1080ti GPU and Intel i7 CPU
A. Lang, S. Vora, H. Caesar, L. Zhou, J. Yang and O. Beijbom: PointPillars: Fast Encoders for Object Detection from Point Clouds. CVPR 2019.
88 STD code 55.04 % 68.33 % 50.85 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Yang, Y. Sun, S. Liu, X. Shen and J. Jia: STD: Sparse-to-Dense 3D Object Detector for Point Cloud. ICCV 2019.
89 Vote3Deep
This method makes use of Velodyne laser scans.
 54.80 % 67.99 % 51.17 % 1.5 s 4 cores @ 2.5 Ghz (C/C++)
M. Engelcke, D. Rao, D. Zeng Wang, C. Hay Tong and I. Posner: Vote3Deep: Fast Object Detection in 3D Point Clouds Using Efficient Convolutional Neural Networks. ArXiv e-prints 2016.
90 yolo4 54.30 % 73.16 % 49.46 % 0.02 s 1 core @ 2.5 Ghz (Python)
91 epBRM
This method makes use of Velodyne laser scans.
 code 54.13 % 62.90 % 51.95 % 0.10 s 1 core @ 2.5 Ghz (C/C++)
K. Shin: Improving a Quality of 3D Object Detection by Spatial Transformation Mechanism. arXiv preprint arXiv:1910.04853 2019.
92 MGACNet 54.13 % 63.54 % 51.79 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
93 PointPainting
This method makes use of Velodyne laser scans.
 53.76 % 61.86 % 50.61 % 0.4 s GPU @ 2.5 Ghz (Python + C/C++)
S. Vora, A. Lang, B. Helou and O. Beijbom: PointPainting: Sequential Fusion for 3D Object Detection. CVPR 2020.
94 PDV2 53.54 % 65.59 % 47.65 % 3.7 s 1 core @ 3.0 Ghz Matlab (C/C++)
J. Shen, X. Zuo, J. Li, W. Yang and H. Ling: A novel pixel neighborhood differential statistic feature for pedestrian and face detection . Pattern Recognition 2017.
95 3DBN_2 53.26 % 63.82 % 50.76 % 0.12 s 1 core @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
96 TAFT 53.15 % 67.62 % 47.08 % 0.2 s 1 core @ 2.5 Ghz (C/C++)
J. Shen, X. Zuo, W. Yang, D. Prokhorov, X. Mei and H. Ling: Differential Features for Pedestrian Detection: A Taylor Series Perspective. IEEE Transactions on Intelligent Transportation Systems 2018.
97 MTMono3d 52.96 % 69.01 % 46.18 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
98 pAUCEnsT 52.88 % 65.84 % 46.97 % 60 s 1 core @ 2.5 Ghz (Matlab + C/C++)
S. Paisitkriangkrai, C. Shen and A. Hengel: Pedestrian Detection with Spatially Pooled Features and Structured Ensemble Learning. arXiv 2014.
99 yolo4_5l code 52.74 % 71.89 % 47.90 % 0.02 s 1 core @ 2.5 Ghz (Python + C/C++)
100 deprecated 52.59 % 60.42 % 50.61 % 0.06 s 1 core @ 2.5 Ghz (C/C++)
101 PFF3D
This method makes use of Velodyne laser scans.
 52.53 % 62.12 % 50.27 % 0.05 s GPU @ 3.0 Ghz (Python + C/C++)
102 ResNet-RRC 52.09 % 66.44 % 47.51 % 0.11 s GPU @ 1.5 Ghz (Python + C/C++)
103 SF
This method uses stereo information.
This method makes use of Velodyne laser scans.
 51.83 % 67.73 % 47.45 % 0.5 s 1 core @ 2.5 Ghz (Python)
ERROR: Wrong syntax in BIBTEX file.
104 Disp R-CNN
This method uses stereo information.
 code 51.36 % 68.93 % 46.79 % 0.42 s GPU @ 2.5 Ghz (Python + C/C++)
J. Sun, L. Chen, Y. Xie, S. Zhang, Q. Jiang, X. Zhou and H. Bao: Disp R-CNN: Stereo 3D Object Detection via Shape Prior Guided Instance Disparity Estimation. CVPR 2020.
105 TBD 51.31 % 61.14 % 47.82 % 0.05 s GPU @ 2.5 Ghz (Python)
106 Disp R-CNN (velo)
This method uses stereo information.
 code 51.31 % 68.84 % 46.80 % 0.42 s GPU @ 2.5 Ghz (Python + C/C++)
J. Sun, L. Chen, Y. Xie, S. Zhang, Q. Jiang, X. Zhou and H. Bao: Disp R-CNN: Stereo 3D Object Detection via Shape Prior Guided Instance Disparity Estimation. CVPR 2020.
107 Shift R-CNN (mono) code 51.30 % 70.86 % 46.37 % 0.25 s GPU @ 1.5 Ghz (Python)
A. Naiden, V. Paunescu, G. Kim, B. Jeon and M. Leordeanu: Shift R-CNN: Deep Monocular 3D Object Detection With Closed-form Geometric Constraints. ICIP 2019.
108 ResNet-RRC (pruned) 51.12 % 65.47 % 46.53 % 0.11 s GPU @ 1.5 Ghz (Python + C/C++)
109 PPFNet code 50.52 % 57.82 % 47.44 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
110 PVNet 50.50 % 60.58 % 48.48 % 0,1 s 1 core @ 2.5 Ghz (Python)
111 LZnet 50.26 % 59.61 % 47.15 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
112 SRDL
This method uses stereo information.
This method makes use of Velodyne laser scans.
 49.72 % 59.18 % 47.82 % 0.15 s GPU @ 2.5 Ghz (Python + C/C++)
113 SCNet
This method makes use of Velodyne laser scans.
 49.61 % 60.95 % 46.91 % 0.04 s GPU @ 3.0 Ghz (Python)
Z. Wang, H. Fu, L. Wang, L. Xiao and B. Dai: SCNet: Subdivision Coding Network for Object Detection Based on 3D Point Cloud. IEEE Access 2019.
114 yolo_depth 49.47 % 67.23 % 44.99 % 0.07 s GPU @ 2.5 Ghz (Python)
115 MMLab-PointRCNN
This method makes use of Velodyne laser scans.
 code 49.41 % 58.93 % 46.44 % 0.1 s GPU @ 2.5 Ghz (Python + C/C++)
S. Shi, X. Wang and H. Li: Pointrcnn: 3d object proposal generation and detection from point cloud. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2019.
116 VOXEL_FPN_HR 49.09 % 60.28 % 45.47 % 0.12 s 8 cores @ 2.5 Ghz (Python)
ERROR: Wrong syntax in BIBTEX file.
117 SS3D_HW 49.01 % 64.67 % 42.86 % 0.4 s GPU @ 2.5 Ghz (Python)
118 Center3D 48.76 % 67.15 % 44.05 % 0.05 s GPU @ 3.5 Ghz (Python)
119 ACFD
This method makes use of Velodyne laser scans.
 code 48.63 % 61.62 % 44.15 % 0.2 s 4 cores @ >3.5 Ghz (C/C++)
M. Dimitrievski, P. Veelaert and W. Philips: Semantically aware multilateral filter for depth upsampling in automotive LiDAR point clouds. IEEE Intelligent Vehicles Symposium, IV 2017, Los Angeles, CA, USA, June 11-14, 2017 2017.
120 R-CNN 48.57 % 62.88 % 43.05 % 4 s GPU @ 3.3 Ghz (C/C++)
J. Hosang, M. Omran, R. Benenson and B. Schiele: Taking a Deeper Look at Pedestrians. arXiv 2015.
121 IGRP+ 48.46 % 59.37 % 44.82 % 0.18 s 1 core @ 2.5 Ghz (Python + C/C++)
122 yolo_rgb 48.45 % 64.50 % 43.95 % 0.07 s GPU @ 2.5 Ghz (Python)
123 EACV 46.88 % 62.21 % 42.46 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
124 PBASN code 46.75 % 54.38 % 44.58 % NA s GPU @ 2.5 Ghz (Python)
125 HR-SECOND code 46.69 % 58.68 % 42.93 % 0.11 s 1 core @ 2.5 Ghz (Python + C/C++)
126 NLK-3D 46.33 % 59.46 % 43.88 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
127 SS3D 45.79 % 61.58 % 41.14 % 48 ms Tesla V100 (Python)
E. Jörgensen, C. Zach and F. Kahl: Monocular 3D Object Detection and Box Fitting Trained End-to-End Using Intersection-over-Union Loss. CoRR 2019.
128 ACF 45.67 % 59.81 % 40.88 % 1 s 1 core @ 3.5 Ghz (Matlab + C/C++)
P. Doll\'ar, R. Appel, S. Belongie and P. Perona: Fast Feature Pyramids for Object Detection. PAMI 2014.
129 NLK-ALL code 45.27 % 56.86 % 41.08 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
130 NL_M3D 45.03 % 58.46 % 39.22 % 0.2 s 1 core @ 2.5 Ghz (Python)
131 Fusion-DPM
This method makes use of Velodyne laser scans.
 code 44.99 % 58.93 % 40.19 % ~ 30 s 1 core @ 3.5 Ghz (Matlab + C/C++)
C. Premebida, J. Carreira, J. Batista and U. Nunes: Pedestrian Detection Combining RGB and Dense LIDAR Data. IROS 2014.
132 ACF-MR 44.79 % 58.29 % 39.94 % 0.6 s 1 core @ 3.5 Ghz (C/C++)
R. Rajaram, E. Ohn-Bar and M. Trivedi: Looking at Pedestrians at Different Scales: A Multi-resolution Approach and Evaluations. T-ITS 2016.
133 yyyyolo 44.55 % 60.74 % 39.96 % 0.01 s 1 core @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
134 HA-SSVM 43.87 % 58.76 % 38.81 % 21 s 1 core @ >3.5 Ghz (Matlab + C/C++)
J. Xu, S. Ramos, D. Vázquez and A. López: Hierarchical Adaptive Structural SVM for Domain Adaptation. IJCV 2016.
135 FCY
This method makes use of Velodyne laser scans.
 43.87 % 56.43 % 39.87 % 0.02 s GPU @ 2.5 Ghz (Python)
136 AB3DMOT
This method makes use of Velodyne laser scans.
This is an online method (no batch processing).
 code 43.86 % 54.55 % 40.99 % 0.0047s 1 core @ 2.5 Ghz (python)
X. Weng and K. Kitani: A Baseline for 3D Multi-Object Tracking. arXiv:1907.03961 2019.
137 D4LCN code 43.50 % 59.55 % 37.12 % 0.2 s GPU @ 2.5 Ghz (Python + C/C++)
M. Ding, Y. Huo, H. Yi, Z. Wang, J. Shi, Z. Lu and P. Luo: Learning Depth-Guided Convolutions for Monocular 3D Object Detection. CVPR 2020.
138 DAMNET code 43.42 % 56.05 % 41.16 % 1 s 1 core @ 2.5 Ghz (C/C++)
139 Pointpillar_TV 43.29 % 53.06 % 41.14 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
140 DPM-VOC+VP 43.26 % 59.21 % 38.12 % 8 s 1 core @ 2.5 Ghz (C/C++)
B. Pepik, M. Stark, P. Gehler and B. Schiele: Multi-view and 3D Deformable Part Models. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) 2015.
141 ACF-SC 42.97 % 53.30 % 38.12 % <0.3 s 1 core @ >3.5 Ghz (Matlab + C/C++)
C. Cadena, A. Dick and I. Reid: A Fast, Modular Scene Understanding System using Context-Aware Object Detection. Robotics and Automation (ICRA), 2015 IEEE International Conference on 2015.
142 SparVox3D 42.73 % 55.22 % 38.08 % 0.05 s GPU @ 2.0 Ghz (Python)
143 SquaresICF code 42.61 % 57.08 % 37.85 % 1 s GPU @ >3.5 Ghz (C/C++)
R. Benenson, M. Mathias, T. Tuytelaars and L. Gool: Seeking the strongest rigid detector. CVPR 2013.
144 CG-Stereo
This method uses stereo information.
 42.54 % 54.64 % 38.45 % 0.57 s GeForce RTX 2080 Ti
C. Li, J. Ku and S. Waslander: Confidence Guided Stereo 3D Object Detection with Split Depth Estimation. IROS 2020.
145 DP3D 42.33 % 57.82 % 38.11 % 0.05 s GPU @ 1.5 Ghz (Python + C/C++)
146 BirdNet+
This method makes use of Velodyne laser scans.
 code 41.97 % 51.38 % 40.15 % 0.1 s Titan Xp (PyTorch)
A. Barrera, C. Guindel, J. Beltrán and F. García: BirdNet+: End-to-End 3D Object Detection in LiDAR Bird's Eye View. arXiv:2003.04188 [cs.CV] 2020.
147 DP3D 41.71 % 55.28 % 35.73 % 0.07 s GPU @ 1.5 Ghz (Python + C/C++)
148 MMRetina
This method uses stereo information.
This method uses optical flow information.
This method makes use of Velodyne laser scans.
 41.63 % 59.63 % 36.97 % 0.38 s GPU @ 2.5 Ghz (Python)
149 CSW3D
This method makes use of Velodyne laser scans.
 41.50 % 53.76 % 37.25 % 0.03 s 4 cores @ 2.5 Ghz (C/C++)
J. Hu, T. Wu, H. Fu, Z. Wang and K. Ding: Cascaded Sliding Window Based Real-Time 3D Region Proposal for Pedestrian Detection. ROBIO 2019.
150 M3D-RPN code 41.46 % 56.64 % 37.31 % 0.16 s GPU @ 1.5 Ghz (Python)
G. Brazil and X. Liu: M3D-RPN: Monocular 3D Region Proposal Network for Object Detection . ICCV 2019 .
151 Stereo3D
This method uses stereo information.
 41.46 % 56.20 % 37.07 % 0.1 s GPU 1080Ti
152 MP-Mono 41.04 % 56.05 % 36.99 % 0.16 s GPU @ 2.5 Ghz (Python)
153 SubCat 40.50 % 53.75 % 35.66 % 1.2 s 6 cores @ 2.5 Ghz (Matlab + C/C++)
E. Ohn-Bar and M. Trivedi: Fast and Robust Object Detection Using Visual Subcategories. Computer Vision and Pattern Recognition Workshops Mobile Vision 2014.
154 vxnet 40.10 % 55.53 % 35.98 % 1 s 1 core @ 2.5 Ghz (C/C++)
155 DSGN
This method uses stereo information.
 code 39.93 % 49.28 % 38.13 % 0.67 s NVIDIA Tesla V100
Y. Chen, S. Liu, X. Shen and J. Jia: DSGN: Deep Stereo Geometry Network for 3D Object Detection. CVPR 2020.
156 RT3D-GMP
This method uses stereo information.
 39.83 % 55.56 % 35.18 % 0.06 s GPU @ 2.5 Ghz (Python + C/C++)
157 SparsePool code 39.59 % 50.81 % 35.91 % 0.13 s 8 cores @ 2.5 Ghz (Python)
Z. Wang, W. Zhan and M. Tomizuka: Fusing bird view lidar point cloud and front view camera image for deep object detection. arXiv preprint arXiv:1711.06703 2017.
158 SparsePool code 39.43 % 50.94 % 35.78 % 0.13 s 8 cores @ 2.5 Ghz (Python)
Z. Wang, W. Zhan and M. Tomizuka: Fusing bird view lidar point cloud and front view camera image for deep object detection. arXiv preprint arXiv:1711.06703 2017.
159 AVOD
This method makes use of Velodyne laser scans.
 code 39.43 % 50.90 % 35.75 % 0.08 s Titan X (pascal)
J. Ku, M. Mozifian, J. Lee, A. Harakeh and S. Waslander: Joint 3D Proposal Generation and Object Detection from View Aggregation. IROS 2018.
160 PG-MonoNet 39.38 % 48.57 % 35.43 % 0.19 s GPU @ 2.5 Ghz (Python)
161 ACF 39.12 % 48.42 % 35.03 % 0.2 s 1 core @ >3.5 Ghz (Matlab + C/C++)
P. Doll\'ar, R. Appel, S. Belongie and P. Perona: Fast Feature Pyramids for Object Detection. PAMI 2014.
P. Doll\'ar: Piotr's Image and Video Matlab Toolbox (PMT). .
162 PB3D
This method uses stereo information.
 38.62 % 50.26 % 34.87 % 0.42 s 1 core @ 2.5 Ghz (C/C++)
163 LSVM-MDPM-sv 37.26 % 50.74 % 33.13 % 10 s 4 cores @ 3.0 Ghz (C/C++)
P. Felzenszwalb, R. Girshick, D. McAllester and D. Ramanan: Object Detection with Discriminatively Trained Part-Based Models. PAMI 2010.
A. Geiger, C. Wojek and R. Urtasun: Joint 3D Estimation of Objects and Scene Layout. NIPS 2011.
164 multi-task CNN 37.00 % 49.38 % 33.46 % 25.1 ms GPU @ 2.0 Ghz (Python)
M. Oeljeklaus, F. Hoffmann and T. Bertram: A Fast Multi-Task CNN for Spatial Understanding of Traffic Scenes. IEEE Intelligent Transportation Systems Conference 2018.
165 Complexer-YOLO
This method makes use of Velodyne laser scans.
 36.45 % 42.16 % 32.91 % 0.06 s GPU @ 3.5 Ghz (C/C++)
M. Simon, K. Amende, A. Kraus, J. Honer, T. Samann, H. Kaulbersch, S. Milz and H. Michael Gross: Complexer-YOLO: Real-Time 3D Object Detection and Tracking on Semantic Point Clouds. The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops 2019.
166 LAPNet 36.22 % 48.96 % 32.27 % 0.03 s 1 core @ 2.5 Ghz (Python)
167 LSVM-MDPM-us code 35.92 % 48.73 % 31.70 % 10 s 4 cores @ 3.0 Ghz (C/C++)
P. Felzenszwalb, R. Girshick, D. McAllester and D. Ramanan: Object Detection with Discriminatively Trained Part-Based Models. PAMI 2010.
168 Vote3D
This method makes use of Velodyne laser scans.
 33.04 % 42.66 % 30.59 % 0.5 s 4 cores @ 2.8 Ghz (C/C++)
D. Wang and I. Posner: Voting for Voting in Online Point Cloud Object Detection. Proceedings of Robotics: Science and Systems 2015.
169 I2BEV 32.42 % 46.35 % 29.98 % 0.63 s GPU @ 2.5 Ghz (Python)
170 OC Stereo
This method uses stereo information.
 code 30.79 % 43.50 % 28.40 % 0.35 s 1 core @ 2.5 Ghz (Python + C/C++)
A. Pon, J. Ku, C. Li and S. Waslander: Object-Centric Stereo Matching for 3D Object Detection. ICRA 2020.
171 mBoW
This method makes use of Velodyne laser scans.
 30.26 % 41.52 % 26.34 % 10 s 1 core @ 2.5 Ghz (C/C++)
J. Behley, V. Steinhage and A. Cremers: Laser-based Segment Classification Using a Mixture of Bag-of-Words. Proc. of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2013.
172 BirdNet
This method makes use of Velodyne laser scans.
 30.07 % 36.82 % 28.40 % 0.11 s Titan Xp (Caffe)
J. Beltrán, C. Guindel, F. Moreno, D. Cruzado, F. García and A. Escalera: BirdNet: A 3D Object Detection Framework from LiDAR Information. 2018 21st International Conference on Intelligent Transportation Systems (ITSC) 2018.
173 RT3DStereo
This method uses stereo information.
 29.30 % 41.12 % 25.25 % 0.08 s GPU @ 2.5 Ghz (C/C++)
H. Königshof, N. Salscheider and C. Stiller: Realtime 3D Object Detection for Automated Driving Using Stereo Vision and Semantic Information. Proc. IEEE Intl. Conf. Intelligent Transportation Systems 2019.
174 DPM-C8B1
This method uses stereo information.
 25.34 % 36.40 % 22.00 % 15 s 4 cores @ 2.5 Ghz (Matlab + C/C++)
J. Yebes, L. Bergasa and M. García-Garrido: Visual Object Recognition with 3D-Aware Features in KITTI Urban Scenes. Sensors 2015.
J. Yebes, L. Bergasa, R. Arroyo and A. Lázaro: Supervised learning and evaluation of KITTI's cars detector with DPM. IV 2014.
175 RefinedMPL 20.81 % 30.41 % 18.72 % 0.15 s GPU @ 2.5 Ghz (Python + C/C++)
J. Vianney, S. Aich and B. Liu: RefinedMPL: Refined Monocular PseudoLiDAR for 3D Object Detection in Autonomous Driving. arXiv preprint arXiv:1911.09712 2019.
176 TopNet-Retina
This method makes use of Velodyne laser scans.
 16.45 % 22.37 % 15.43 % 52ms GeForce 1080Ti (tensorflow-gpu, v1.12)
S. Wirges, T. Fischer, C. Stiller and J. Frias: Object Detection and Classification in Occupancy Grid Maps Using Deep Convolutional Networks. 2018 21st International Conference on Intelligent Transportation Systems (ITSC) 2018.
177 TopNet-HighRes
This method makes use of Velodyne laser scans.
 15.28 % 21.22 % 13.89 % 101ms NVIDIA GeForce 1080 Ti (tensorflow-gpu)
S. Wirges, T. Fischer, C. Stiller and J. Frias: Object Detection and Classification in Occupancy Grid Maps Using Deep Convolutional Networks. 2018 21st International Conference on Intelligent Transportation Systems (ITSC) 2018.
178 YOLOv2 code 11.46 % 15.37 % 9.67 % 0.02 s GPU @ 3.5 Ghz (C/C++)
J. Redmon, S. Divvala, R. Girshick and A. Farhadi: You only look once: Unified, real-time object detection. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2016.
J. Redmon and A. Farhadi: YOLO9000: Better, Faster, Stronger. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2017.
179 TopNet-UncEst
This method makes use of Velodyne laser scans.
 8.58 % 13.00 % 7.38 % 0.09 s NVIDIA GeForce 1080 Ti (tensorflow-gpu)
S. Wirges, M. Braun, M. Lauer and C. Stiller: Capturing Object Detection Uncertainty in Multi-Layer Grid Maps. 2019.
180 BIP-HETERO 7.05 % 8.51 % 6.30 % ~2 s 1 core @ 2.5 Ghz (C/C++)
A. Mekonnen, F. Lerasle, A. Herbulot and C. Briand: People Detection with Heterogeneous Features and Explicit Optimization on Computation Time. Pattern Recognition (ICPR), 2014 22nd International Conference on 2014.
181 CBNet 1.33 % 1.03 % 1.41 % 1 s 4 cores @ 2.5 Ghz (Python)
182 TopNet-DecayRate
This method makes use of Velodyne laser scans.
 0.01 % 0.01 % 0.01 % 92 ms NVIDIA GeForce 1080 Ti (tensorflow-gpu)
S. Wirges, T. Fischer, C. Stiller and J. Frias: Object Detection and Classification in Occupancy Grid Maps Using Deep Convolutional Networks. 2018 21st International Conference on Intelligent Transportation Systems (ITSC) 2018.
183 UM3D_TUM 0.00 % 0.00 % 0.00 % 0.05 s 1 core @ 2.5 Ghz (Python)
Table as LaTeX | Only published Methods

Cyclist


Method Setting Code Moderate Easy Hard Runtime Environment
1 HRI-MSP-L
This method makes use of Velodyne laser scans.
 83.08 % 92.12 % 75.62 % 0.07 s 1 core @ 2.5 Ghz (C/C++)
2 Deformable PV-RCNN
This method makes use of Velodyne laser scans.
 code 80.57 % 88.65 % 74.81 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
P. Bhattacharyya and K. Czarnecki: Deformable PV-RCNN: Improving 3D Object Detection with Learned Deformations. ECCV 2020 Perception for Autonomous Driving Workshop.
3 MMLab PV-RCNN
This method makes use of Velodyne laser scans.
 code 80.42 % 86.62 % 73.64 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
S. Shi, C. Guo, L. Jiang, Z. Wang, J. Shi, X. Wang and H. Li: PV-RCNN: Point-Voxel Feature Set Abstraction for 3D Object Detection. CVPR 2020.
4 Noah CV Lab - SSL 79.10 % 86.71 % 69.66 % 0.1 s GPU @ 2.5 Ghz (Python)
5 HotSpotNet 78.81 % 86.06 % 71.74 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
Q. Chen, L. Sun, Z. Wang, K. Jia and A. Yuille: object as hotspots. Proceedings of the European Conference on Computer Vision (ECCV) 2020.
6 MMLab-PartA^2
This method makes use of Velodyne laser scans.
 code 78.29 % 88.90 % 71.19 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
S. Shi, Z. Wang, J. Shi, X. Wang and H. Li: From Points to Parts: 3D Object Detection from Point Cloud with Part-aware and Part-aggregation Network. IEEE Transactions on Pattern Analysis and Machine Intelligence 2020.
7 IC-PVRCNN 78.27 % 88.66 % 72.05 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
8 F-ConvNet
This method makes use of Velodyne laser scans.
 code 78.05 % 86.75 % 68.12 % 0.47 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Wang and K. Jia: Frustum ConvNet: Sliding Frustums to Aggregate Local Point-Wise Features for Amodal 3D Object Detection. IROS 2019.
9 PointPainting
This method makes use of Velodyne laser scans.
 78.04 % 87.70 % 69.27 % 0.4 s GPU @ 2.5 Ghz (Python + C/C++)
S. Vora, A. Lang, B. Helou and O. Beijbom: PointPainting: Sequential Fusion for 3D Object Detection. CVPR 2020.
10 MMCOM 77.43 % 85.83 % 68.34 % 0.04 s 1 core @ 2.5 Ghz (Python)
11 TBD 77.34 % 87.15 % 70.53 % 0.05 s GPU @ 2.5 Ghz (Python)
12 CVIS-DF3D_v2 76.98 % 86.55 % 70.29 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
13 RRC code 76.81 % 86.81 % 66.59 % 3.6 s GPU @ 2.5 Ghz (C/C++)
J. Ren, X. Chen, J. Liu, W. Sun, J. Pang, Q. Yan, Y. Tai and L. Xu: Accurate Single Stage Detector Using Recurrent Rolling Convolution. CVPR 2017.
14 KNN-GCNN 76.52 % 88.83 % 69.82 % 0.4 s 1 core @ 2.5 Ghz (Python + C/C++)
15 IC-SECOND 76.47 % 85.99 % 70.98 % 0.06 s 1 core @ 2.5 Ghz (C/C++)
16 deprecated 76.14 % 84.03 % 70.81 % 0.06 s 1 core @ 2.5 Ghz (C/C++)
17 HWFD 75.54 % 85.88 % 66.85 % 0.21 s one 1080Ti
18 MVX-Net++ 75.41 % 86.78 % 68.49 % 0.15 s 1 core @ 2.5 Ghz (C/C++)
19 MS-CNN code 75.30 % 84.88 % 65.27 % 0.4 s GPU @ 2.5 Ghz (C/C++)
Z. Cai, Q. Fan, R. Feris and N. Vasconcelos: A Unified Multi-scale Deep Convolutional Neural Network for Fast Object Detection. ECCV 2016.
20 VOXEL_FPN_HR 75.24 % 87.73 % 68.60 % 0.12 s 8 cores @ 2.5 Ghz (Python)
ERROR: Wrong syntax in BIBTEX file.
21 RethinkDet3D 75.22 % 89.04 % 66.47 % 0.15 s 1 core @ 2.5 Ghz (Python)
22 TuSimple code 75.22 % 83.68 % 65.22 % 1.6 s GPU @ 2.5 Ghz (Python + C/C++)
F. Yang, W. Choi and Y. Lin: Exploit all the layers: Fast and accurate cnn object detector with scale dependent pooling and cascaded rejection classifiers. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2016.
K. He, X. Zhang, S. Ren and J. Sun: Deep residual learning for image recognition. Proceedings of the IEEE conference on computer vision and pattern recognition 2016.
23 Point-GNN
This method makes use of Velodyne laser scans.
 code 75.08 % 85.75 % 68.69 % 0.6 s GPU @ 2.5 Ghz (Python)
W. Shi and R. Rajkumar: Point-GNN: Graph Neural Network for 3D Object Detection in a Point Cloud. CVPR 2020.
24 PP-3D 75.08 % 85.75 % 68.69 % 0.1 s 1 core @ 2.5 Ghz (Python)
25 NLK-ALL code 74.83 % 87.37 % 68.05 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
26 Deep3DBox 74.78 % 84.36 % 64.05 % 1.5 s GPU @ 2.5 Ghz (C/C++)
A. Mousavian, D. Anguelov, J. Flynn and J. Kosecka: 3D Bounding Box Estimation Using Deep Learning and Geometry. CVPR 2017.
27 SRDL
This method uses stereo information.
This method makes use of Velodyne laser scans.
 74.68 % 88.17 % 68.29 % 0.15 s GPU @ 2.5 Ghz (Python + C/C++)
28 PPBA 74.46 % 86.45 % 69.15 % NA s GPU @ 2.5 Ghz (Python)
29 TBU 74.46 % 86.45 % 69.15 % NA s GPU @ 2.5 Ghz (Python)
30 3DBN_2 74.34 % 88.48 % 67.66 % 0.12 s 1 core @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
31 3DSSD code 74.12 % 87.09 % 67.67 % 0.04 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Yang, Y. Sun, S. Liu and J. Jia: 3DSSD: Point-based 3D Single Stage Object Detector. CVPR 2020.
32 LZnet 73.93 % 87.74 % 68.54 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
33 SDP+RPN 73.85 % 82.59 % 64.87 % 0.4 s GPU @ 2.5 Ghz (Python + C/C++)
F. Yang, W. Choi and Y. Lin: Exploit All the Layers: Fast and Accurate CNN Object Detector with Scale Dependent Pooling and Cascaded Rejection Classifiers. Proceedings of the IEEE International Conference on Computer Vision and Pattern Recognition 2016.
S. Ren, K. He, R. Girshick and J. Sun: Faster R-CNN: Towards real-time object detection with region proposal networks. Advances in Neural Information Processing Systems 2015.
34 Baseline of CA RCNN 73.69 % 85.39 % 66.94 % 0.1 s GPU @ 2.5 Ghz (Python)
35 CVIS-DF3D 73.69 % 85.39 % 66.94 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
36 SVGA-Net
This method makes use of Velodyne laser scans.
 73.68 % 85.44 % 66.94 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
37 MGACNet 73.66 % 85.45 % 67.10 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
38 dgist_multiDetNet 73.57 % 87.95 % 64.65 % 0.08 s GPU Titanx Pascal (Python)
39 sensekitti code 73.48 % 82.90 % 64.03 % 4.5 s GPU @ 2.5 Ghz (Python + C/C++)
B. Yang, J. Yan, Z. Lei and S. Li: Craft Objects from Images. CVPR 2016.
40 CA-RCNN 73.45 % 85.13 % 66.08 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
41 MMLab-PointRCNN
This method makes use of Velodyne laser scans.
 code 73.42 % 86.21 % 66.45 % 0.1 s GPU @ 2.5 Ghz (Python + C/C++)
S. Shi, X. Wang and H. Li: Pointrcnn: 3d object proposal generation and detection from point cloud. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2019.
42 F-PointNet
This method makes use of Velodyne laser scans.
 code 73.16 % 86.86 % 65.21 % 0.17 s GPU @ 3.0 Ghz (Python)
C. Qi, W. Liu, C. Wu, H. Su and L. Guibas: Frustum PointNets for 3D Object Detection from RGB-D Data. arXiv preprint arXiv:1711.08488 2017.
43 HR-SECOND code 72.77 % 84.21 % 66.25 % 0.11 s 1 core @ 2.5 Ghz (Python + C/C++)
44 MonoPSR code 72.08 % 82.06 % 62.43 % 0.2 s GPU @ 3.5 Ghz (Python)
J. Ku*, A. Pon* and S. Waslander: Monocular 3D Object Detection Leveraging Accurate Proposals and Shape Reconstruction. CVPR 2019.
45 ARPNET 71.95 % 84.96 % 65.21 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
Y. Ye, C. Zhang and X. Hao: ARPNET: attention region proposal network for 3D object detection. Science China Information Sciences 2019.
46 SubCNN 71.72 % 79.36 % 62.74 % 2 s GPU @ 3.5 Ghz (Python + C/C++)
Y. Xiang, W. Choi, Y. Lin and S. Savarese: Subcategory-aware Convolutional Neural Networks for Object Proposals and Detection. IEEE Winter Conference on Applications of Computer Vision (WACV) 2017.
47 STD code 71.63 % 83.99 % 64.92 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Yang, Y. Sun, S. Liu, X. Shen and J. Jia: STD: Sparse-to-Dense 3D Object Detector for Point Cloud. ICCV 2019.
48 PiP 71.52 % 82.97 % 65.52 % 0.033 s 1 core @ 2.5 Ghz (Python)
49 PVNet 71.10 % 83.89 % 65.08 % 0,1 s 1 core @ 2.5 Ghz (Python)
50 Multi-task DG 70.98 % 80.96 % 62.18 % 0.06 s GPU @ 2.5 Ghz (Python)
51 Faster RCNN + Gr + A 70.78 % 83.99 % 63.36 % 1.29 s GPU @ 2.5 Ghz (Python)
52 NLK-3D 70.55 % 85.92 % 63.76 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
53 PBASN code 70.21 % 83.96 % 65.10 % NA s GPU @ 2.5 Ghz (Python)
54 AB3DMOT
This method makes use of Velodyne laser scans.
This is an online method (no batch processing).
 code 70.18 % 82.86 % 63.55 % 0.0047s 1 core @ 2.5 Ghz (Python)
X. Weng and K. Kitani: A Baseline for 3D Multi-Object Tracking. arXiv:1907.03961 2019.
55 AP-RCNN 70.11 % 86.24 % 63.26 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
56 FCY
This method makes use of Velodyne laser scans.
 70.05 % 83.02 % 63.63 % 0.02 s GPU @ 2.5 Ghz (Python)
57 PointPillars
This method makes use of Velodyne laser scans.
 code 68.98 % 83.97 % 62.17 % 16 ms 1080ti GPU and Intel i7 CPU
A. Lang, S. Vora, H. Caesar, L. Zhou, J. Yang and O. Beijbom: PointPillars: Fast Encoders for Object Detection from Point Clouds. CVPR 2019.
58 CentrNet-FG 68.88 % 83.29 % 61.47 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
59 Vote3Deep
This method makes use of Velodyne laser scans.
 68.82 % 78.41 % 62.50 % 1.5 s 4 cores @ 2.5 Ghz (C/C++)
M. Engelcke, D. Rao, D. Zeng Wang, C. Hay Tong and I. Posner: Vote3Deep: Fast Object Detection in 3D Point Clouds Using Efficient Convolutional Neural Networks. ArXiv e-prints 2016.
60 3DOP
This method uses stereo information.
 code 68.71 % 80.52 % 61.07 % 3s GPU @ 2.5 Ghz (Matlab + C/C++)
X. Chen, K. Kundu, Y. Zhu, A. Berneshawi, H. Ma, S. Fidler and R. Urtasun: 3D Object Proposals for Accurate Object Class Detection. NIPS 2015.
61 Pose-RCNN 68.40 % 81.53 % 59.43 % 2 s >8 cores @ 2.5 Ghz (Python)
M. Braun, Q. Rao, Y. Wang and F. Flohr: Pose-RCNN: Joint object detection and pose estimation using 3D object proposals. Intelligent Transportation Systems (ITSC), 2016 IEEE 19th International Conference on 2016.
62 TANet code 68.20 % 82.24 % 62.13 % 0.035s GPU @ 2.5 Ghz (Python + C/C++)
Z. Liu, X. Zhao, T. Huang, R. Hu, Y. Zhou and X. Bai: TANet: Robust 3D Object Detection from Point Clouds with Triple Attention. AAAI 2020.
63 Faster RCNN + G 68.09 % 83.51 % 60.60 % 1.1 s GPU @ 2.5 Ghz (Python)
64 VICNet 68.07 % 86.63 % 61.03 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
65 Faster RCNN + A 67.84 % 82.06 % 60.52 % 0.19 s GPU @ 2.5 Ghz (Python)
66 IVA code 67.57 % 78.48 % 58.83 % 0.4 s GPU @ 2.5 Ghz (C/C++)
Y. Zhu, J. Wang, C. Zhao, H. Guo and H. Lu: Scale-adaptive Deconvolutional Regression Network for Pedestrian Detection. ACCV 2016.
S. Ren, K. He, R. Girshick and J. Sun: Faster R-CNN: Towards real-time object detection with region proposal networks. Advances in neural information processing systems 2015.
67 IGRP+ 67.56 % 83.94 % 61.15 % 0.18 s 1 core @ 2.5 Ghz (Python + C/C++)
68 DeepStereoOP 67.22 % 79.35 % 58.60 % 3.4 s GPU @ 3.5 Ghz (Matlab + C/C++)
C. Pham and J. Jeon: Robust Object Proposals Re-ranking for Object Detection in Autonomous Driving Using Convolutional Neural Networks. Signal Processing: Image Communiation 2017.
69 Faster RCNN + A 67.15 % 83.77 % 59.85 % 0.19 s GPU @ 2.5 Ghz (Python + C/C++)
70 FII-CenterNet 66.54 % 79.04 % 57.76 % 0.09 s GPU @ 2.5 Ghz (Python)
71 epBRM
This method makes use of Velodyne laser scans.
 code 66.51 % 79.65 % 60.31 % 0.10 s 1 core @ 2.5 Ghz (C/C++)
K. Shin: Improving a Quality of 3D Object Detection by Spatial Transformation Mechanism. arXiv preprint arXiv:1910.04853 2019.
72 PFF3D
This method makes use of Velodyne laser scans.
 66.25 % 79.44 % 60.11 % 0.05 s GPU @ 3.0 Ghz (Python + C/C++)
73 Pointpillar_TV 66.20 % 79.86 % 59.73 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
74 HR-faster-rcnn 65.53 % 83.49 % 58.13 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
75 Mono3D code 65.15 % 77.19 % 57.88 % 4.2 s GPU @ 2.5 Ghz (Matlab + C/C++)
X. Chen, K. Kundu, Z. Zhang, H. Ma, S. Fidler and R. Urtasun: Monocular 3D Object Detection for Autonomous Driving. CVPR 2016.
76 Simple3D Net 64.77 % 79.60 % 58.48 % 0.02 s GPU @ 2.5 Ghz (Python)
77 Mono3CN 63.29 % 81.46 % 56.27 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
78 Faster R-CNN code 62.86 % 72.40 % 54.97 % 2 s GPU @ 3.5 Ghz (Python + C/C++)
S. Ren, K. He, R. Girshick and J. Sun: Faster R-CNN: Towards Real- Time Object Detection with Region Proposal Networks. NIPS 2015.
79 SCNet
This method makes use of Velodyne laser scans.
 62.50 % 78.48 % 56.34 % 0.04 s GPU @ 3.0 Ghz (Python)
Z. Wang, H. Fu, L. Wang, L. Xiao and B. Dai: SCNet: Subdivision Coding Network for Object Detection Based on 3D Point Cloud. IEEE Access 2019.
80 UDI-mono3D 62.26 % 79.20 % 53.95 % 0.05 s 1 core @ 2.5 Ghz (Python)
81 AVOD-FPN
This method makes use of Velodyne laser scans.
 code 60.79 % 70.38 % 55.37 % 0.1 s Titan X (Pascal)
J. Ku, M. Mozifian, J. Lee, A. Harakeh and S. Waslander: Joint 3D Proposal Generation and Object Detection from View Aggregation. IROS 2018.
82 SDP+CRC (ft) 60.72 % 75.63 % 53.00 % 0.6 s GPU @ 2.5 Ghz (C/C++)
F. Yang, W. Choi and Y. Lin: Exploit All the Layers: Fast and Accurate CNN Object Detector with Scale Dependent Pooling and Cascaded Rejection Classifiers. Proceedings of the IEEE International Conference on Computer Vision and Pattern Recognition 2016.
83 Complexer-YOLO
This method makes use of Velodyne laser scans.
 59.78 % 66.94 % 55.63 % 0.06 s GPU @ 3.5 Ghz (C/C++)
M. Simon, K. Amende, A. Kraus, J. Honer, T. Samann, H. Kaulbersch, S. Milz and H. Michael Gross: Complexer-YOLO: Real-Time 3D Object Detection and Tracking on Semantic Point Clouds. The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops 2019.
84 UDI-mono3D 59.44 % 77.70 % 51.49 % 0.05 s 1 core @ 2.5 Ghz (Python)
85 Regionlets 58.52 % 71.12 % 50.83 % 1 s >8 cores @ 2.5 Ghz (C/C++)
X. Wang, M. Yang, S. Zhu and Y. Lin: Regionlets for Generic Object Detection. T-PAMI 2015.
W. Zou, X. Wang, M. Sun and Y. Lin: Generic Object Detection with Dense Neural Patterns and Regionlets. British Machine Vision Conference 2014.
C. Long, X. Wang, G. Hua, M. Yang and Y. Lin: Accurate Object Detection with Location Relaxation and Regionlets Relocalization. Asian Conference on Computer Vision 2014.
86 DAM 58.41 % 76.09 % 49.93 % 1 s GPU @ 2.5 Ghz (Python)
87 GA_rpn500 57.82 % 76.06 % 49.00 % 1 s 1 core @ 2.5 Ghz (Python)
88 GA2500 57.82 % 76.06 % 48.99 % 0.2 s 1 core @ 2.5 Ghz (Python)
89 PMN 57.22 % 74.63 % 49.82 % 0.2 s 1 core @ 2.5 Ghz (Python)
90 GA_FULLDATA 57.20 % 75.50 % 50.26 % 1 s 4 cores @ 2.5 Ghz (Python)
91 FRCNN+Or code 57.01 % 70.99 % 50.14 % 0.09 s Titan Xp GPU
C. Guindel, D. Martin and J. Armingol: Fast Joint Object Detection and Viewpoint Estimation for Traffic Scene Understanding. IEEE Intelligent Transportation Systems Magazine 2018.
C. Guindel, D. Martin and J. Armingol: Joint Object Detection and Viewpoint Estimation using CNN features. IEEE International Conference on Vehicular Electronics and Safety (ICVES) 2017.
92 MonoPair 56.37 % 74.77 % 48.37 % 0.06 s GPU @ 2.5 Ghz (Python + C/C++)
Y. Chen, L. Tai, K. Sun and M. Li: MonoPair: Monocular 3D Object Detection Using Pairwise Spatial Relationships. The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2020.
93 GA_BALANCE 56.07 % 78.33 % 49.02 % 1 s 1 core @ 2.5 Ghz (Python)
94 MLOD
This method makes use of Velodyne laser scans.
 code 56.04 % 75.35 % 49.11 % 0.12 s GPU @ 1.5 Ghz (Python)
J. Deng and K. Czarnecki: MLOD: A multi-view 3D object detection based on robust feature fusion method. arXiv preprint arXiv:1909.04163 2019.
95 bigger_ga 55.66 % 73.05 % 47.31 % 1 s 1 core @ 2.5 Ghz (Python)
96 yolo4_5l 55.42 % 75.21 % 48.57 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
97 BirdNet+
This method makes use of Velodyne laser scans.
 code 54.61 % 74.97 % 50.29 % 0.1 s Titan Xp (PyTorch)
A. Barrera, C. Guindel, J. Beltrán and F. García: BirdNet+: End-to-End 3D Object Detection in LiDAR Bird's Eye View. arXiv:2003.04188 [cs.CV] 2020.
98 NL_M3D 53.51 % 71.09 % 47.07 % 0.2 s 1 core @ 2.5 Ghz (Python)
99 CRCNNA 53.41 % 69.81 % 46.29 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
100 DAMNET code 52.93 % 71.01 % 48.56 % 1 s 1 core @ 2.5 Ghz (C/C++)
101 AVOD
This method makes use of Velodyne laser scans.
 code 52.60 % 66.45 % 46.39 % 0.08 s Titan X (pascal)
J. Ku, M. Mozifian, J. Lee, A. Harakeh and S. Waslander: Joint 3D Proposal Generation and Object Detection from View Aggregation. IROS 2018.
102 GAFM 51.40 % 73.43 % 44.61 % 0.5 s 1 core @ 2.5 Ghz (Python)
103 JSU-NET 51.10 % 72.92 % 44.26 % 0.1 s 1 core @ 2.5 Ghz (Python)
104 yolo4 50.62 % 71.71 % 44.18 % 0.02 s 1 core @ 2.5 Ghz (Python)
105 CDI3D 50.29 % 63.72 % 43.95 % 0.03 s GPU @ 2.5 Ghz (Python)
106 ga50 49.02 % 70.25 % 42.52 % 1 s 1 core @ 2.5 Ghz (Python)
107 Mag 48.91 % 68.41 % 42.87 % 0.07 s 1 core @ 2.5 Ghz (C/C++)
108 yolo4 48.67 % 67.33 % 43.00 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
109 CG-Stereo
This method uses stereo information.
 48.46 % 69.98 % 42.41 % 0.57 s GeForce RTX 2080 Ti
C. Li, J. Ku and S. Waslander: Confidence Guided Stereo 3D Object Detection with Split Depth Estimation. IROS 2020.
110 MP-Mono 48.38 % 65.58 % 39.97 % 0.16 s GPU @ 2.5 Ghz (Python)
111 yolo4_5l code 48.38 % 69.14 % 42.16 % 0.02 s 1 core @ 2.5 Ghz (Python + C/C++)
112 MTMono3d 47.71 % 67.12 % 38.84 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
113 BirdNet
This method makes use of Velodyne laser scans.
 47.64 % 64.91 % 44.59 % 0.11 s Titan Xp (Caffe)
J. Beltrán, C. Guindel, F. Moreno, D. Cruzado, F. García and A. Escalera: BirdNet: A 3D Object Detection Framework from LiDAR Information. 2018 21st International Conference on Intelligent Transportation Systems (ITSC) 2018.
114 SS3D_HW 45.53 % 61.79 % 39.03 % 0.4 s GPU @ 2.5 Ghz (Python)
115 SparsePool code 44.57 % 60.53 % 40.37 % 0.13 s 8 cores @ 2.5 Ghz (Python)
Z. Wang, W. Zhan and M. Tomizuka: Fusing bird view lidar point cloud and front view camera image for deep object detection. arXiv preprint arXiv:1711.06703 2017.
116 Scan_YOLO 43.39 % 64.82 % 37.77 % 0.1 s 4 cores @ 3.0 Ghz (Python)
117 ResNet-RRC (pruned) 43.35 % 58.81 % 37.68 % 0.11 s GPU @ 1.5 Ghz (Python + C/C++)
118 Shift R-CNN (mono) code 42.96 % 63.24 % 38.22 % 0.25 s GPU @ 1.5 Ghz (Python)
A. Naiden, V. Paunescu, G. Kim, B. Jeon and M. Leordeanu: Shift R-CNN: Deep Monocular 3D Object Detection With Closed-form Geometric Constraints. ICIP 2019.
119 ResNet-RRC 42.88 % 58.72 % 37.74 % 0.11 s GPU @ 1.5 Ghz (Python + C/C++)
120 D4LCN code 42.86 % 65.29 % 36.29 % 0.2 s GPU @ 2.5 Ghz (Python + C/C++)
M. Ding, Y. Huo, H. Yi, Z. Wang, J. Shi, Z. Lu and P. Luo: Learning Depth-Guided Convolutions for Monocular 3D Object Detection. CVPR 2020.
121 Disp R-CNN (velo)
This method uses stereo information.
 code 42.25 % 58.27 % 36.90 % 0.42 s GPU @ 2.5 Ghz (Python + C/C++)
J. Sun, L. Chen, Y. Xie, S. Zhang, Q. Jiang, X. Zhou and H. Bao: Disp R-CNN: Stereo 3D Object Detection via Shape Prior Guided Instance Disparity Estimation. CVPR 2020.
122 Disp R-CNN
This method uses stereo information.
 code 42.23 % 58.26 % 36.88 % 0.42 s GPU @ 2.5 Ghz (Python + C/C++)
J. Sun, L. Chen, Y. Xie, S. Zhang, Q. Jiang, X. Zhou and H. Bao: Disp R-CNN: Stereo 3D Object Detection via Shape Prior Guided Instance Disparity Estimation. CVPR 2020.
123 vxnet 41.64 % 60.52 % 36.89 % 1 s 1 core @ 2.5 Ghz (C/C++)
124 yolo_rgb 41.59 % 62.22 % 37.32 % 0.07 s GPU @ 2.5 Ghz (Python)
125 M3D-RPN code 41.54 % 61.54 % 35.23 % 0.16 s GPU @ 1.5 Ghz (Python)
G. Brazil and X. Liu: M3D-RPN: Monocular 3D Region Proposal Network for Object Detection . ICCV 2019 .
126 Center3D 40.99 % 65.34 % 36.50 % 0.05 s GPU @ 3.5 Ghz (Python)
127 MV-RGBD-RF
This method makes use of Velodyne laser scans.
 40.94 % 51.10 % 34.83 % 4 s 4 cores @ 2.5 Ghz (C/C++)
A. Gonzalez, D. Vazquez, A. Lopez and J. Amores: On-Board Object Detection: Multicue, Multimodal, and Multiview Random Forest of Local Experts.. IEEE Trans. on Cybernetics 2016.
A. Gonzalez, G. Villalonga, J. Xu, D. Vazquez, J. Amores and A. Lopez: Multiview Random Forest of Local Experts Combining RGB and LIDAR data for Pedestrian Detection. IEEE Intelligent Vehicles Symposium (IV) 2015.
128 DP3D 37.13 % 53.50 % 32.82 % 0.05 s GPU @ 1.5 Ghz (Python + C/C++)
129 yolo_depth 36.89 % 50.88 % 32.64 % 0.07 s GPU @ 2.5 Ghz (Python)
130 SparsePool code 36.26 % 44.21 % 32.57 % 0.13 s 8 cores @ 2.5 Ghz (Python)
Z. Wang, W. Zhan and M. Tomizuka: Fusing bird view lidar point cloud and front view camera image for deep object detection. arXiv preprint arXiv:1711.06703 2017.
131 PG-MonoNet 36.09 % 47.28 % 32.15 % 0.19 s GPU @ 2.5 Ghz (Python)
132 DP3D 36.05 % 52.18 % 30.19 % 0.07 s GPU @ 1.5 Ghz (Python + C/C++)
133 LAPNet 35.62 % 53.46 % 29.27 % 0.03 s 1 core @ 2.5 Ghz (Python)
134 SS3D 35.48 % 52.97 % 31.07 % 48 ms Tesla V100 (Python)
E. Jörgensen, C. Zach and F. Kahl: Monocular 3D Object Detection and Box Fitting Trained End-to-End Using Intersection-over-Union Loss. CoRR 2019.
135 DSGN
This method uses stereo information.
 code 35.15 % 49.10 % 31.41 % 0.67 s NVIDIA Tesla V100
Y. Chen, S. Liu, X. Shen and J. Jia: DSGN: Deep Stereo Geometry Network for 3D Object Detection. CVPR 2020.
136 pAUCEnsT 34.90 % 50.51 % 30.35 % 60 s 1 core @ 2.5 Ghz (Matlab + C/C++)
S. Paisitkriangkrai, C. Shen and A. Hengel: Pedestrian Detection with Spatially Pooled Features and Structured Ensemble Learning. arXiv 2014.
137 EACV 34.34 % 50.21 % 30.13 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
138 TopNet-Retina
This method makes use of Velodyne laser scans.
 31.98 % 47.51 % 29.84 % 52ms GeForce 1080Ti (tensorflow-gpu, v1.12)
S. Wirges, T. Fischer, C. Stiller and J. Frias: Object Detection and Classification in Occupancy Grid Maps Using Deep Convolutional Networks. 2018 21st International Conference on Intelligent Transportation Systems (ITSC) 2018.
139 PB3D
This method uses stereo information.
 28.78 % 45.05 % 25.66 % 0.42 s 1 core @ 2.5 Ghz (C/C++)
140 OC Stereo
This method uses stereo information.
 code 28.76 % 43.18 % 24.80 % 0.35 s 1 core @ 2.5 Ghz (Python + C/C++)
A. Pon, J. Ku, C. Li and S. Waslander: Object-Centric Stereo Matching for 3D Object Detection. ICRA 2020.
141 MMRetina
This method uses stereo information.
This method uses optical flow information.
This method makes use of Velodyne laser scans.
 28.00 % 43.71 % 24.62 % 0.38 s GPU @ 2.5 Ghz (Python)
142 Vote3D
This method makes use of Velodyne laser scans.
 27.99 % 39.81 % 25.19 % 0.5 s 4 cores @ 2.8 Ghz (C/C++)
D. Wang and I. Posner: Voting for Voting in Online Point Cloud Object Detection. Proceedings of Robotics: Science and Systems 2015.
143 LSVM-MDPM-us code 27.81 % 37.66 % 24.83 % 10 s 4 cores @ 3.0 Ghz (C/C++)
P. Felzenszwalb, R. Girshick, D. McAllester and D. Ramanan: Object Detection with Discriminatively Trained Part-Based Models. PAMI 2010.
144 DPM-VOC+VP 27.73 % 41.58 % 24.61 % 8 s 1 core @ 2.5 Ghz (C/C++)
B. Pepik, M. Stark, P. Gehler and B. Schiele: Multi-view and 3D Deformable Part Models. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) 2015.
145 RefinedMPL 27.17 % 44.47 % 22.84 % 0.15 s GPU @ 2.5 Ghz (Python + C/C++)
J. Vianney, S. Aich and B. Liu: RefinedMPL: Refined Monocular PseudoLiDAR for 3D Object Detection in Autonomous Driving. arXiv preprint arXiv:1911.09712 2019.
146 I2BEV 27.13 % 40.03 % 23.23 % 0.63 s GPU @ 2.5 Ghz (Python)
147 LSVM-MDPM-sv 26.05 % 35.70 % 23.56 % 10 s 4 cores @ 3.0 Ghz (C/C++)
P. Felzenszwalb, R. Girshick, D. McAllester and D. Ramanan: Object Detection with Discriminatively Trained Part-Based Models. PAMI 2010.
A. Geiger, C. Wojek and R. Urtasun: Joint 3D Estimation of Objects and Scene Layout. NIPS 2011.
148 DPM-C8B1
This method uses stereo information.
 25.57 % 41.47 % 21.93 % 15 s 4 cores @ 2.5 Ghz (Matlab + C/C++)
J. Yebes, L. Bergasa and M. García-Garrido: Visual Object Recognition with 3D-Aware Features in KITTI Urban Scenes. Sensors 2015.
J. Yebes, L. Bergasa, R. Arroyo and A. Lázaro: Supervised learning and evaluation of KITTI's cars detector with DPM. IV 2014.
149 BdCost+DA+BB+MS 25.52 % 33.92 % 21.14 % TBD s 4 cores @ 2.5 Ghz (C/C++)
150 RT3D-GMP
This method uses stereo information.
 22.90 % 33.64 % 19.87 % 0.06 s GPU @ 2.5 Ghz (Python + C/C++)
151 BdCost+DA+BB 20.00 % 26.87 % 16.76 % TBD s 4 cores @ 2.5 Ghz (C/C++)
152 mBoW
This method makes use of Velodyne laser scans.
 17.63 % 26.66 % 16.02 % 10 s 1 core @ 2.5 Ghz (C/C++)
J. Behley, V. Steinhage and A. Cremers: Laser-based Segment Classification Using a Mixture of Bag-of-Words. Proc. of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2013.
153 TopNet-HighRes
This method makes use of Velodyne laser scans.
 13.98 % 22.86 % 14.52 % 101ms NVIDIA GeForce 1080 Ti (tensorflow-gpu)
S. Wirges, T. Fischer, C. Stiller and J. Frias: Object Detection and Classification in Occupancy Grid Maps Using Deep Convolutional Networks. 2018 21st International Conference on Intelligent Transportation Systems (ITSC) 2018.
154 RT3DStereo
This method uses stereo information.
 12.96 % 19.58 % 11.47 % 0.08 s GPU @ 2.5 Ghz (C/C++)
H. Königshof, N. Salscheider and C. Stiller: Realtime 3D Object Detection for Automated Driving Using Stereo Vision and Semantic Information. Proc. IEEE Intl. Conf. Intelligent Transportation Systems 2019.
155 yyyyolo 12.52 % 16.29 % 11.07 % 0.01 s 1 core @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
156 TopNet-UncEst
This method makes use of Velodyne laser scans.
 12.00 % 18.14 % 11.85 % 0.09 s NVIDIA GeForce 1080 Ti (tensorflow-gpu)
S. Wirges, M. Braun, M. Lauer and C. Stiller: Capturing Object Detection Uncertainty in Multi-Layer Grid Maps. 2019.
157 CBNet 0.39 % 0.24 % 0.44 % 1 s 4 cores @ 2.5 Ghz (Python)
158 YOLOv2 code 0.06 % 0.15 % 0.07 % 0.02 s GPU @ 3.5 Ghz (C/C++)
J. Redmon, S. Divvala, R. Girshick and A. Farhadi: You only look once: Unified, real-time object detection. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2016.
J. Redmon and A. Farhadi: YOLO9000: Better, Faster, Stronger. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2017.
159 TopNet-DecayRate
This method makes use of Velodyne laser scans.
 0.04 % 0.00 % 0.04 % 92 ms NVIDIA GeForce 1080 Ti (tensorflow-gpu)
S. Wirges, T. Fischer, C. Stiller and J. Frias: Object Detection and Classification in Occupancy Grid Maps Using Deep Convolutional Networks. 2018 21st International Conference on Intelligent Transportation Systems (ITSC) 2018.
160 UM3D_TUM 0.00 % 0.00 % 0.00 % 0.05 s 1 core @ 2.5 Ghz (Python)
Table as LaTeX | Only published Methods

Object Detection and Orientation Estimation Evaluation

Cars


Method Setting Code Moderate Easy Hard Runtime Environment
1 CLOCs_PVCas 95.79 % 96.74 % 90.81 % 0.1 s 1 core @ 2.5 Ghz (Python)
S. Pang, D. Morris and H. Radha: CLOCs: Camera-LiDAR Object Candidates Fusion for 3D Object Detection . 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2020.
2 BorderAtt 95.44 % 96.37 % 90.31 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
3 HUAWEI Octopus 95.40 % 96.29 % 92.67 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
4 SPANet 95.03 % 96.31 % 89.99 % 0.06 s 1 core @ 2.5 Ghz (C/C++)
5 Voxel R-CNN 94.96 % 96.47 % 92.24 % 0.04 s GPU @ 3.0 Ghz (C/C++)
6 MMLab PV-RCNN
This method makes use of Velodyne laser scans.
 code 94.57 % 98.15 % 91.85 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
S. Shi, C. Guo, L. Jiang, Z. Wang, J. Shi, X. Wang and H. Li: PV-RCNN: Point-Voxel Feature Set Abstraction for 3D Object Detection. CVPR 2020.
7 IC-SECOND 94.55 % 95.78 % 91.67 % 0.06 s 1 core @ 2.5 Ghz (C/C++)
8 PC-RGNN 94.55 % 95.79 % 92.03 % 0.1 s GPU @ 2.5 Ghz (Python + C/C++)
9 Deformable PV-RCNN
This method makes use of Velodyne laser scans.
 code 94.52 % 95.84 % 91.93 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
P. Bhattacharyya and K. Czarnecki: Deformable PV-RCNN: Improving 3D Object Detection with Learned Deformations. ECCV 2020 Perception for Autonomous Driving Workshop.
10 ReFineNet 94.49 % 95.74 % 91.97 % 0.08 s 1 core @ 2.5 Ghz (Python)
11 nonet 94.48 % 95.85 % 91.64 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
12 HyBrid Feature Det 94.47 % 95.75 % 92.00 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
13 MVRA + I-FRCNN+ 94.46 % 95.66 % 81.74 % 0.18 s GPU @ 2.5 Ghz (Python)
H. Choi, H. Kang and Y. Hyun: Multi-View Reprojection Architecture for Orientation Estimation. The IEEE International Conference on Computer Vision (ICCV) Workshops 2019.
14 MGACNet 94.44 % 95.33 % 91.55 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
15 RangeRCNN-LV 94.37 % 95.92 % 91.85 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
16 OneCoLab SicNet V2 94.35 % 95.64 % 91.92 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
17 AP-RCNN 94.31 % 96.10 % 89.92 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
18 IC-PVRCNN 94.29 % 95.88 % 91.77 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
19 SERCNN
This method makes use of Velodyne laser scans.
 94.24 % 96.31 % 89.71 % 0.1 s 1 core @ 2.5 Ghz (Python)
D. Zhou, J. Fang, X. Song, L. Liu, J. Yin, Y. Dai, H. Li and R. Yang: Joint 3D Instance Segmentation and Object Detection for Autonomous Driving. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition 2020.
20 CVRS_PF 94.23 % 95.55 % 91.21 % 0.09 s 1 core @ 2.5 Ghz (C/C++)
21 EPNet code 94.22 % 96.13 % 89.68 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
T. Huang, Z. Liu, X. Chen and X. Bai: EPNet: Enhancing Point Features with Image Semantics for 3D Object Detection. ECCV 2020.
22 D3D 94.18 % 95.22 % 89.14 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
23 CVIS-DF3D_v2 94.16 % 95.68 % 91.45 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
24 SVGA-Net
This method makes use of Velodyne laser scans.
 94.08 % 95.83 % 91.55 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
25 Baseline of CA RCNN 94.07 % 95.82 % 91.54 % 0.1 s GPU @ 2.5 Ghz (Python)
26 CVIS-DF3D 94.07 % 95.82 % 91.54 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
27 CA-RCNN 94.04 % 95.78 % 91.27 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
28 tbd code 94.03 % 95.66 % 91.20 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
29 RangeRCNN
This method makes use of Velodyne laser scans.
 93.90 % 95.47 % 91.53 % 0.06 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Liang, M. Zhang, Z. Zhang, X. Zhao and S. Pu: RangeRCNN: Towards Fast and Accurate 3D Object Detection with Range Image Representation. arXiv preprint arXiv:2009.00206 2020.
30 HRI-MSP-L
This method makes use of Velodyne laser scans.
 93.82 % 95.50 % 91.27 % 0.07 s 1 core @ 2.5 Ghz (C/C++)
31 AF_V1 93.77 % 94.45 % 86.25 % 0.1 s 1 core @ 2.5 Ghz (Python)
32 Patches - EMP
This method makes use of Velodyne laser scans.
 93.58 % 97.88 % 90.31 % 0.5 s GPU @ 2.5 Ghz (Python)
J. Lehner, A. Mitterecker, T. Adler, M. Hofmarcher, B. Nessler and S. Hochreiter: Patch Refinement: Localized 3D Object Detection. arXiv preprint arXiv:1910.04093 2019.
33 Associate-3Ddet_v2 93.46 % 96.66 % 88.20 % 0.04 s 1 core @ 2.5 Ghz (Python)
34 VAL 93.45 % 96.83 % 83.46 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
35 OAP 93.35 % 96.56 % 85.69 % 0.06 s 1 core @ 2.5 Ghz (C/C++)
36 CLOCs_PointCas 93.34 % 96.66 % 85.87 % 0.1 s GPU @ 2.5 Ghz (Python)
37 CIA-SSD
This method makes use of Velodyne laser scans.
 93.34 % 96.65 % 85.76 % 0.03 s 1 core @ 2.5 Ghz (Python + C/C++)
38 Deep MANTA 93.31 % 98.83 % 82.95 % 0.7 s GPU @ 2.5 Ghz (Python + C/C++)
F. Chabot, M. Chaouch, J. Rabarisoa, C. Teulière and T. Chateau: Deep MANTA: A Coarse-to-fine Many-Task Network for joint 2D and 3D vehicle analysis from monocular image. CVPR 2017.
39 CIA-SSD v2
This method makes use of Velodyne laser scans.
 93.22 % 96.53 % 87.81 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
40 deprecated 93.22 % 96.75 % 85.64 % deprecated deprecated
41 CBi-GNN 93.16 % 98.70 % 87.97 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
42 PointRes
This method makes use of Velodyne laser scans.
 93.15 % 96.57 % 90.04 % 0.013 s 1 core @ 2.5 Ghz (Python + C/C++)
43 AIMC-RUC 93.14 % 96.64 % 87.92 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
44 CJJ 93.14 % 96.59 % 90.16 % 0.04 s 1 core @ 2.5 Ghz (C/C++)
45 FCY
This method makes use of Velodyne laser scans.
 93.08 % 96.51 % 87.90 % 0.02 s GPU @ 2.5 Ghz (Python)
46 ISF 93.05 % 96.52 % 90.08 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
47 MVX-Net++ 92.93 % 96.16 % 87.69 % 0.15 s 1 core @ 2.5 Ghz (C/C++)
48 scssd-normal(0.3) 92.88 % 96.44 % 87.67 % 0.05 s GPU @ 2.5 Ghz (Python)
P. An, J. Liang, J. Ma, K. Yu and B. Fang: SCSSD for multi sensor fusion 3d object detection method. Submitted to Information Science 2020.
49 Cas-SSD 92.83 % 96.38 % 87.64 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
50 FLID 92.77 % 95.64 % 85.00 % 0.04 s GPU @ 2.5 Ghz (Python)
51 HotSpotNet 92.74 % 96.20 % 89.68 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
Q. Chen, L. Sun, Z. Wang, K. Jia and A. Yuille: object as hotspots. Proceedings of the European Conference on Computer Vision (ECCV) 2020.
52 scssd-normal(0.4) 92.71 % 96.24 % 87.55 % 0.05 s 1 core @ 2.5 Ghz (Python)
P. An, J. Liang, J. Ma, K. Yu and B. Fang: SCSSD for multi sensor fusion 3d object detection method. Submitted to Information Science 2020.
53 IGRP 92.66 % 96.27 % 87.63 % 0.18 s 1 core @ 2.5 Ghz (Python + C/C++)
54 deprecated 92.66 % 95.52 % 91.39 % 0.06 s 1 core @ 2.5 Ghz (C/C++)
55 SARPNET 92.58 % 95.82 % 87.33 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
Y. Ye, H. Chen, C. Zhang, X. Hao and Z. Zhang: SARPNET: Shape Attention Regional Proposal Network for LiDAR-based 3D Object Detection. Neurocomputing 2019.
56 Patches
This method makes use of Velodyne laser scans.
 92.57 % 96.31 % 87.41 % 0.15 s GPU @ 2.0 Ghz
J. Lehner, A. Mitterecker, T. Adler, M. Hofmarcher, B. Nessler and S. Hochreiter: Patch Refinement: Localized 3D Object Detection. arXiv preprint arXiv:1910.04093 2019.
57 R-GCN 92.53 % 96.16 % 87.45 % 0.16 s GPU @ 2.5 Ghz (Python)
J. Zarzar, S. Giancola and B. Ghanem: PointRGCN: Graph Convolution Networks for 3D Vehicles Detection Refinement. ArXiv 2019.
58 PPFNet code 92.52 % 96.30 % 87.44 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
59 PI-RCNN 92.52 % 96.15 % 87.47 % 0.1 s 1 core @ 2.5 Ghz (Python)
L. Xie, C. Xiang, Z. Yu, G. Xu, Z. Yang, D. Cai and X. He: PI-RCNN: An Efficient Multi-sensor 3D Object Detector with Point-based Attentive Cont-conv Fusion Module. AAAI 2020 : The Thirty-Fourth AAAI Conference on Artificial Intelligence 2020.
60 CenterNet3D 92.48 % 95.71 % 89.54 % 0.04 s GPU @ 1.5 Ghz (Python)
G. Wang, B. Tian, Y. Ai, T. Xu, L. Chen and D. Cao: CenterNet3D:An Anchor free Object Detector for Autonomous Driving. 2020.
61 Discrete-PointDet 92.48 % 95.89 % 87.08 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
62 PointPainting
This method makes use of Velodyne laser scans.
 92.43 % 98.36 % 89.49 % 0.4 s GPU @ 2.5 Ghz (Python + C/C++)
S. Vora, A. Lang, B. Helou and O. Beijbom: PointPainting: Sequential Fusion for 3D Object Detection. CVPR 2020.
63 3D IoU-Net 92.42 % 96.31 % 87.60 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
J. Li, S. Luo, Z. Zhu, H. Dai, S. Krylov, Y. Ding and L. Shao: 3D IoU-Net: IoU Guided 3D Object Detector for Point Clouds. arXiv preprint arXiv:2004.04962 2020.
64 CLOCs_SecCas 92.37 % 95.16 % 88.43 % 0.1 s 1 core @ 2.5 Ghz (Python)
S. Pang, D. Morris and H. Radha: CLOCs: Camera-LiDAR Object Candidates Fusion for 3D Object Detection. 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2020.
65 VAR 92.28 % 95.08 % 89.42 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
66 NLK-ALL code 92.27 % 95.49 % 87.34 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
67 OneCoLab SicNet 92.23 % 95.53 % 89.60 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
68 LZY_RCNN 92.18 % 93.57 % 89.61 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
69 SegVoxelNet 92.16 % 95.86 % 86.90 % 0.04 s 1 core @ 2.5 Ghz (Python)
H. Yi, S. Shi, M. Ding, J. Sun, K. Xu, H. Zhou, Z. Wang, S. Li and G. Wang: SegVoxelNet: Exploring Semantic Context and Depth-aware Features for 3D Vehicle Detection from Point Cloud. ICRA 2020.
70 PointRGCN 92.15 % 97.48 % 86.83 % 0.26 s GPU @ V100 (Python)
J. Zarzar, S. Giancola and B. Ghanem: PointRGCN: Graph Convolution Networks for 3D Vehicles Detection Refinement. ArXiv 2019.
71 NLK-3D 92.15 % 95.20 % 87.13 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
72 RethinkDet3D 92.04 % 95.68 % 86.97 % 0.15 s 1 core @ 2.5 Ghz (Python)
73 MDA 92.01 % 94.87 % 89.31 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
74 PVNet 92.00 % 94.82 % 89.08 % 0,1 s 1 core @ 2.5 Ghz (Python)
75 F-ConvNet
This method makes use of Velodyne laser scans.
 code 91.98 % 95.81 % 79.83 % 0.47 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Wang and K. Jia: Frustum ConvNet: Sliding Frustums to Aggregate Local Point-Wise Features for Amodal 3D Object Detection. IROS 2019.
76 TBD 91.97 % 93.46 % 89.36 % 0.05 s GPU @ 2.5 Ghz (Python)
77 PointCSE 91.95 % 95.52 % 86.75 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
78 IE-PointRCNN 91.94 % 96.00 % 86.84 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
79 SRDL
This method uses stereo information.
This method makes use of Velodyne laser scans.
 91.93 % 95.56 % 89.30 % 0.15 s GPU @ 2.5 Ghz (Python + C/C++)
80 AB3DMOT
This method makes use of Velodyne laser scans.
This is an online method (no batch processing).
 code 91.87 % 95.86 % 86.78 % 0.0047s 1 core @ 2.5 Ghz (Python)
X. Weng and K. Kitani: A Baseline for 3D Multi-Object Tracking. arXiv:1907.03961 2019.
81 MMLab-PointRCNN
This method makes use of Velodyne laser scans.
 code 91.77 % 95.90 % 86.92 % 0.1 s GPU @ 2.5 Ghz (Python + C/C++)
S. Shi, X. Wang and H. Li: Pointrcnn: 3d object proposal generation and detection from point cloud. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2019.
82 MMLab-PartA^2
This method makes use of Velodyne laser scans.
 code 91.73 % 95.00 % 88.86 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
S. Shi, Z. Wang, J. Shi, X. Wang and H. Li: From Points to Parts: 3D Object Detection from Point Cloud with Part-aware and Part-aggregation Network. IEEE Transactions on Pattern Analysis and Machine Intelligence 2020.
83 Pointpillar_TV 91.61 % 94.80 % 88.25 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
84 C-GCN 91.57 % 95.63 % 86.13 % 0.147 s GPU @ V100 (Python)
J. Zarzar, S. Giancola and B. Ghanem: PointRGCN: Graph Convolution Networks for 3D Vehicles Detection Refinement. ArXiv 2019.
85 V3D 91.44 % 94.45 % 86.22 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
86 TBD 91.42 % 95.56 % 86.05 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
87 tt code 91.38 % 95.14 % 88.39 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
88 CU-PointRCNN 91.25 % 97.24 % 86.85 % 0.1 s GPU @ 1.5 Ghz (Python + C/C++)
89 RUC 91.25 % 95.01 % 88.14 % 0.12 s 1 core @ 2.5 Ghz (C/C++)
90 deprecated 91.18 % 96.19 % 83.25 % 0.05 s 1 core @ 2.5 Ghz (Python)
91 PFF3D
This method makes use of Velodyne laser scans.
 91.06 % 94.86 % 86.28 % 0.05 s GPU @ 3.0 Ghz (Python + C/C++)
92 3DBN_2 91.05 % 94.89 % 88.42 % 0.12 s 1 core @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
93 VICNet 91.00 % 94.66 % 85.74 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
94 Mono3CN 90.96 % 94.22 % 82.86 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
95 MAFF-Net(DAF-Pillar) 90.78 % 94.17 % 83.17 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
Z. Zhang, Z. Liang, M. Zhang, X. Zhao, Y. Ming, T. Wenming and S. Pu: MAFF-Net: Filter False Positive for 3D Vehicle Detection with Multi-modal Adaptive Feature Fusion. arXiv preprint arXiv:2009.10945 2020.
96 HRI-VoxelFPN 90.76 % 96.35 % 85.37 % 0.02 s GPU @ 2.5 Ghz (Python + C/C++)
H. Kuang, B. Wang, J. An, M. Zhang and Z. Zhang: Voxel-FPN:multi-scale voxel feature aggregation in 3D object detection from point clouds. sensors 2020.
97 SSL-RTM3D 90.70 % 96.34 % 80.72 % 0.03 s 1 core @ 2.5 Ghz (Python)
98 anonymous 90.70 % 96.46 % 82.39 % 1 s 1 core @ 2.5 Ghz (C/C++)
99 PointPillars
This method makes use of Velodyne laser scans.
 code 90.70 % 93.84 % 87.47 % 16 ms 1080ti GPU and Intel i7 CPU
A. Lang, S. Vora, H. Caesar, L. Zhou, J. Yang and O. Beijbom: PointPillars: Fast Encoders for Object Detection from Point Clouds. CVPR 2019.
100 WS3D
This method makes use of Velodyne laser scans.
 90.69 % 94.85 % 85.94 % 0.1 s GPU @ 2.5 Ghz (Python)
Q. Meng, W. Wang, T. Zhou, J. Shen, L. Van Gool and D. Dai: Weakly Supervised 3D Object Detection from Lidar Point Cloud. 2020.
101 DPointNet 90.50 % 93.70 % 85.40 % 0.09 s 1 core @ 2.5 Ghz (C/C++)
102 OACV 90.35 % 93.95 % 81.90 % 0.23 s GPU @ 2.5 Ghz (Python)
103 IGRP+ 90.31 % 96.01 % 87.41 % 0.18 s 1 core @ 2.5 Ghz (Python + C/C++)
104 autonet 90.31 % 93.30 % 87.00 % 0.12 s 1 core @ 2.5 Ghz (C/C++)
105 3D IoU Loss
This method makes use of Velodyne laser scans.
 90.21 % 95.60 % 84.96 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
D. Zhou, J. Fang, X. Song, C. Guan, J. Yin, Y. Dai and R. Yang: IoU Loss for 2D/3D Object Detection. International Conference on 3D Vision (3DV) 2019.
106 Bit 90.19 % 93.42 % 86.48 % 0.11 s 1 core @ 2.5 Ghz (C/C++)
107 ARPNET 90.11 % 93.42 % 82.56 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
Y. Ye, C. Zhang and X. Hao: ARPNET: attention region proposal network for 3D object detection. Science China Information Sciences 2019.
108 TANet code 90.11 % 93.52 % 84.61 % 0.035s GPU @ 2.5 Ghz (Python + C/C++)
Z. Liu, X. Zhao, T. Huang, R. Hu, Y. Zhou and X. Bai: TANet: Robust 3D Object Detection from Point Clouds with Triple Attention. AAAI 2020.
109 EPENet 90.09 % 93.83 % 86.76 % 0.04 s 1 core @ 2.5 Ghz (C/C++)
110 CentrNet-FG 90.04 % 93.51 % 87.02 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
111 OCM3D 90.03 % 94.18 % 83.29 % 0.5 s 1 core @ 2.5 Ghz (C/C++)
112 CG-Stereo
This method uses stereo information.
 89.98 % 96.28 % 82.21 % 0.57 s GeForce RTX 2080 Ti
C. Li, J. Ku and S. Waslander: Confidence Guided Stereo 3D Object Detection with Split Depth Estimation. IROS 2020.
113 Det3D 89.92 % 94.21 % 83.18 % 0.5 s 1 core @ 2.5 Ghz (C/C++)
114 Deep3DBox 89.88 % 94.62 % 76.40 % 1.5 s GPU @ 2.5 Ghz (C/C++)
A. Mousavian, D. Anguelov, J. Flynn and J. Kosecka: 3D Bounding Box Estimation Using Deep Learning and Geometry. CVPR 2017.
115 PointPiallars_SECA 89.86 % 92.96 % 86.46 % 0.06 s 1 core @ 2.5 Ghz (C/C++)
116 VOXEL_FPN_HR 89.81 % 93.52 % 84.59 % 0.12 s 8 cores @ 2.5 Ghz (Python)
ERROR: Wrong syntax in BIBTEX file.
117 BVVF 89.77 % 95.55 % 84.48 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
118 baseline 89.69 % 92.61 % 86.03 % 0.12 s 1 core @ 2.5 Ghz (C/C++)
119 GPP code 89.68 % 93.94 % 80.60 % 0.23 s GPU @ 1.5 Ghz (Python + C/C++)
A. Rangesh and M. Trivedi: Ground plane polling for 6dof pose estimation of objects on the road. arXiv preprint arXiv:1811.06666 2018.
120 SubCNN 89.53 % 94.11 % 79.14 % 2 s GPU @ 3.5 Ghz (Python + C/C++)
Y. Xiang, W. Choi, Y. Lin and S. Savarese: Subcategory-aware Convolutional Neural Networks for Object Proposals and Detection. IEEE Winter Conference on Applications of Computer Vision (WACV) 2017.
121 SCNet
This method makes use of Velodyne laser scans.
 89.36 % 95.23 % 84.03 % 0.04 s GPU @ 3.0 Ghz (Python)
Z. Wang, H. Fu, L. Wang, L. Xiao and B. Dai: SCNet: Subdivision Coding Network for Object Detection Based on 3D Point Cloud. IEEE Access 2019.
122 RUC code 89.26 % 92.28 % 85.38 % 0.12 s 1 core @ 2.5 Ghz (C/C++)
123 AVOD
This method makes use of Velodyne laser scans.
 code 89.22 % 94.98 % 82.14 % 0.08 s Titan X (pascal)
J. Ku, M. Mozifian, J. Lee, A. Harakeh and S. Waslander: Joint 3D Proposal Generation and Object Detection from View Aggregation. IROS 2018.
124 IAFA 89.14 % 92.96 % 79.40 % 0.04 s 1 core @ 2.5 Ghz (C/C++)
125 MCA 88.91 % 92.91 % 79.11 % 0.04 s 1 core @ 2.5 Ghz (C/C++)
126 RUC code 88.90 % 92.68 % 84.04 % 0.12 s 1 core @ 2.5 Ghz (C/C++)
127 AVOD-FPN
This method makes use of Velodyne laser scans.
 code 88.61 % 94.65 % 83.71 % 0.1 s Titan X (Pascal)
J. Ku, M. Mozifian, J. Lee, A. Harakeh and S. Waslander: Joint 3D Proposal Generation and Object Detection from View Aggregation. IROS 2018.
128 SS3D_HW 88.50 % 94.45 % 68.61 % 0.4 s GPU @ 2.5 Ghz (Python)
129 PSMD 88.29 % 93.59 % 75.35 % 0.1 s GPU @ 2.5 Ghz (Python)
130 Prune 88.10 % 93.86 % 80.41 % 0.11 s 1 core @ 2.5 Ghz (C/C++)
131 autoRUC 88.03 % 93.80 % 80.36 % 0.12 s 1 core @ 2.5 Ghz (C/C++)
132 AACL 88.00 % 93.36 % 73.17 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
133 DeepStereoOP 87.81 % 93.68 % 77.60 % 3.4 s GPU @ 3.5 Ghz (Matlab + C/C++)
C. Pham and J. Jeon: Robust Object Proposals Re-ranking for Object Detection in Autonomous Driving Using Convolutional Neural Networks. Signal Processing: Image Communiation 2017.
134 3DBN
This method makes use of Velodyne laser scans.
 87.59 % 93.34 % 79.91 % 0.13s 1080Ti (Python+C/C++)
X. Li, J. Guivant, N. Kwok and Y. Xu: 3D Backbone Network for 3D Object Detection. CoRR 2019.
135 ISF-v2 87.49 % 93.15 % 84.13 % 0.04 s 1 core @ 2.5 Ghz (Python)
136 FQNet 87.49 % 93.66 % 73.61 % 0.5 s 1 core @ 2.5 Ghz (Python)
L. Liu, J. Lu, C. Xu, Q. Tian and J. Zhou: Deep Fitting Degree Scoring Network for Monocular 3D Object Detection. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2019.
137 Shift R-CNN (mono) code 87.47 % 93.75 % 77.19 % 0.25 s GPU @ 1.5 Ghz (Python)
A. Naiden, V. Paunescu, G. Kim, B. Jeon and M. Leordeanu: Shift R-CNN: Deep Monocular 3D Object Detection With Closed-form Geometric Constraints. ICIP 2019.
138 MonoPSR code 87.45 % 93.29 % 72.26 % 0.2 s GPU @ 3.5 Ghz (Python)
J. Ku*, A. Pon* and S. Waslander: Monocular 3D Object Detection Leveraging Accurate Proposals and Shape Reconstruction. CVPR 2019.
139 Mono3D code 87.28 % 93.13 % 77.00 % 4.2 s GPU @ 2.5 Ghz (Matlab + C/C++)
X. Chen, K. Kundu, Z. Zhang, H. Ma, S. Fidler and R. Urtasun: Monocular 3D Object Detection for Autonomous Driving. CVPR 2016.
140 MA 87.08 % 93.12 % 79.50 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
141 SMOKE code 87.02 % 92.94 % 77.12 % 0.03 s GPU @ 2.5 Ghz (Python)
Z. Liu, Z. Wu and R. Tóth: SMOKE: Single-Stage Monocular 3D Object Detection via Keypoint Estimation. 2020.
142 3DOP
This method uses stereo information.
 code 86.93 % 91.31 % 76.72 % 3s GPU @ 2.5 Ghz (Matlab + C/C++)
X. Chen, K. Kundu, Y. Zhu, A. Berneshawi, H. Ma, S. Fidler and R. Urtasun: 3D Object Proposals for Accurate Object Class Detection. NIPS 2015.
143 CDN
This method uses stereo information.
 86.90 % 95.79 % 79.05 % 0.6 s GPU @ 2.5 Ghz (Python)
D. Garg, Y. Wang, B. Hariharan, M. Campbell, K. Weinberger and W. Chao: Wasserstein Distances for Stereo Disparity Estimation. Advances in Neural Information Processing Systems (NeurIPS) 2020.
144 tiny-stereo-v2
This method uses stereo information.
 86.87 % 95.91 % 79.43 % 0.4 s 1 core @ 2.5 Ghz (Python + C/C++)
145 DAMNET code 86.83 % 92.37 % 81.73 % 1 s 1 core @ 2.5 Ghz (C/C++)
146 RTM3D code 86.73 % 91.75 % 77.18 % 0.05 s GPU @ 1.0 Ghz (Python)
P. Li, H. Zhao, P. Liu and F. Cao: RTM3D: Real-time Monocular 3D Detection from Object Keypoints for Autonomous Driving. 2020.
147 IMA 86.71 % 92.51 % 76.95 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
148 voxelrcnn 86.61 % 94.59 % 79.80 % 15 s 1 core @ 2.5 Ghz (C/C++)
149 MonoPair 86.11 % 91.65 % 76.45 % 0.06 s GPU @ 2.5 Ghz (Python + C/C++)
Y. Chen, L. Tai, K. Sun and M. Li: MonoPair: Monocular 3D Object Detection Using Pairwise Spatial Relationships. The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2020.
150 DSGN
This method uses stereo information.
 code 86.03 % 95.42 % 78.27 % 0.67 s NVIDIA Tesla V100
Y. Chen, S. Liu, X. Shen and J. Jia: DSGN: Deep Stereo Geometry Network for 3D Object Detection. CVPR 2020.
151 tiny-stereo-v1
This method uses stereo information.
 85.77 % 95.09 % 77.89 % 0.3 s GPU @ 2.5 Ghz (Python + C/C++)
152 UDI-mono3D 85.76 % 92.25 % 76.23 % 0.05 s 1 core @ 2.5 Ghz (Python)
153 NL_M3D 85.32 % 90.88 % 70.87 % 0.2 s 1 core @ 2.5 Ghz (Python)
154 StereoFENet
This method uses stereo information.
 85.14 % 91.28 % 76.80 % 0.15 s 1 core @ 3.5 Ghz (Python)
W. Bao, B. Xu and Z. Chen: MonoFENet: Monocular 3D Object Detection with Feature Enhancement Networks. IEEE Transactions on Image Processing 2019.
155 PB3D
This method uses stereo information.
 84.75 % 95.15 % 75.34 % 0.42 s 1 core @ 2.5 Ghz (C/C++)
156 PL++ (SDN+GDC)
This method uses stereo information.
This method makes use of Velodyne laser scans.
 code 84.42 % 94.83 % 76.95 % 0.6 s GPU @ 2.5 Ghz (C/C++)
Y. You, Y. Wang, W. Chao, D. Garg, G. Pleiss, B. Hariharan, M. Campbell and K. Weinberger: Pseudo-LiDAR++: Accurate Depth for 3D Object Detection in Autonomous Driving. International Conference on Learning Representations 2020.
157 SS3D 84.38 % 92.57 % 69.82 % 48 ms Tesla V100 (Python)
E. Jörgensen, C. Zach and F. Kahl: Monocular 3D Object Detection and Box Fitting Trained End-to-End Using Intersection-over-Union Loss. CoRR 2019.
158 IDA-3D
This method uses stereo information.
 84.32 % 92.63 % 73.98 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
159 CDN-PL++
This method uses stereo information.
 84.21 % 94.45 % 76.69 % 0.4 s GPU @ 2.5 Ghz (C/C++)
D. Garg, Y. Wang, B. Hariharan, M. Campbell, K. Weinberger and W. Chao: Wasserstein Distances for Stereo Disparity Estimation. arXiv preprint arXiv:2007.03085 2020.
160 UDI-mono3D 84.20 % 91.88 % 75.38 % 0.05 s 1 core @ 2.5 Ghz (Python)
161 MonoFENet 84.09 % 91.42 % 75.93 % 0.15 s 1 core @ 3.5 Ghz (Python)
W. Bao, B. Xu and Z. Chen: MonoFENet: Monocular 3D Object Detection with Feature Enhancement Networks. IEEE Transactions on Image Processing 2019.
162 Complexer-YOLO
This method makes use of Velodyne laser scans.
 83.89 % 91.77 % 79.24 % 0.06 s GPU @ 3.5 Ghz (C/C++)
M. Simon, K. Amende, A. Kraus, J. Honer, T. Samann, H. Kaulbersch, S. Milz and H. Michael Gross: Complexer-YOLO: Real-Time 3D Object Detection and Tracking on Semantic Point Clouds. The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops 2019.
163 ZoomNet
This method uses stereo information.
 code 83.79 % 94.14 % 68.78 % 0.3 s 1 core @ 2.5 Ghz (C/C++)
L. Z. Xu: ZoomNet: Part-Aware Adaptive Zooming Neural Network for 3D Object Detection. Proceedings of the AAAI Conference on Artificial Intelligence 2020.
164 seivl 83.38 % 90.32 % 81.41 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
165 MP-Mono 83.09 % 91.04 % 64.30 % 0.16 s GPU @ 2.5 Ghz (Python)
166 M3D-RPN code 82.81 % 88.38 % 67.08 % 0.16 s GPU @ 1.5 Ghz (Python)
G. Brazil and X. Liu: M3D-RPN: Monocular 3D Region Proposal Network for Object Detection . ICCV 2019 .
167 MTMono3d 82.65 % 90.34 % 74.98 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
168 RAR-Net 82.63 % 88.40 % 66.90 % 0.5 s 1 core @ 2.5 Ghz (C/C++)
169 Center3D 82.51 % 93.10 % 70.79 % 0.05 s GPU @ 3.5 Ghz (Python)
170 SSL-RTM3D Res18 82.43 % 93.13 % 72.47 % 0.02 s GPU @ 2.5 Ghz (Python)
171 vxnet 82.40 % 87.78 % 61.28 % 1 s 1 core @ 2.5 Ghz (C/C++)
172 ASOD 82.13 % 93.56 % 67.32 % 0.28 s GPU @ 2.5 Ghz (Python)
173 D4LCN code 82.08 % 90.01 % 63.98 % 0.2 s GPU @ 2.5 Ghz (Python + C/C++)
M. Ding, Y. Huo, H. Yi, Z. Wang, J. Shi, Z. Lu and P. Luo: Learning Depth-Guided Convolutions for Monocular 3D Object Detection. CVPR 2020.
174 deprecated 81.99 % 92.07 % 67.48 % 1 core @ 2.5 Ghz (C/C++)
175 S3D 81.93 % 91.59 % 67.43 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
176 Pseudo-LiDAR++
This method uses stereo information.
 code 81.87 % 94.14 % 74.29 % 0.4 s GPU @ 2.5 Ghz (Python)
Y. You, Y. Wang, W. Chao, D. Garg, G. Pleiss, B. Hariharan, M. Campbell and K. Weinberger: Pseudo-LiDAR++: Accurate Depth for 3D Object Detection in Autonomous Driving. International Conference on Learning Representations 2020.
177 LNET 81.81 % 91.36 % 67.33 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
178 Disp R-CNN
This method uses stereo information.
 code 81.70 % 93.02 % 67.16 % 0.42 s GPU @ 2.5 Ghz (Python + C/C++)
J. Sun, L. Chen, Y. Xie, S. Zhang, Q. Jiang, X. Zhou and H. Bao: Disp R-CNN: Stereo 3D Object Detection via Shape Prior Guided Instance Disparity Estimation. CVPR 2020.
179 Disp R-CNN (velo)
This method uses stereo information.
 code 81.67 % 92.86 % 67.22 % 0.42 s GPU @ 2.5 Ghz (Python + C/C++)
J. Sun, L. Chen, Y. Xie, S. Zhang, Q. Jiang, X. Zhou and H. Bao: Disp R-CNN: Stereo 3D Object Detection via Shape Prior Guided Instance Disparity Estimation. CVPR 2020.
180 LAPNet 81.63 % 90.16 % 63.47 % 0.03 s 1 core @ 2.5 Ghz (Python)
181 Pseudo-LiDAR E2E
This method uses stereo information.
 81.56 % 93.74 % 74.23 % 0.4 s GPU @ 2.5 Ghz (Python)
182 HR-SECOND code 81.23 % 88.32 % 74.89 % 0.11 s 1 core @ 2.5 Ghz (Python + C/C++)
183 BS3D 81.22 % 94.66 % 68.39 % 22 ms Titan Xp
N. Gählert, J. Wan, M. Weber, J. Zöllner, U. Franke and J. Denzler: Beyond Bounding Boxes: Using Bounding Shapes for Real-Time 3D Vehicle Detection from Monocular RGB Images. 2019 IEEE Intelligent Vehicles Symposium (IV) 2019.
184 DP3D 81.07 % 87.49 % 65.12 % 0.05 s GPU @ 1.5 Ghz (Python + C/C++)
185 LCD3D 81.01 % 91.20 % 64.29 % 0.03 s GPU @ 2.5 Ghz (Python)
186 Stereo3D
This method uses stereo information.
 80.88 % 93.65 % 61.17 % 0.1 s GPU 1080Ti
187 DP3D 80.87 % 87.58 % 64.88 % 0.07 s GPU @ 1.5 Ghz (Python + C/C++)
188 FRCNN+Or code 80.57 % 91.50 % 67.49 % 0.09 s Titan Xp GPU
C. Guindel, D. Martin and J. Armingol: Fast Joint Object Detection and Viewpoint Estimation for Traffic Scene Understanding. IEEE Intelligent Transportation Systems Magazine 2018.
C. Guindel, D. Martin and J. Armingol: Joint Object Detection and Viewpoint Estimation using CNN features. IEEE International Conference on Vehicular Electronics and Safety (ICVES) 2017.
189 UM3D_TUM 80.15 % 92.80 % 65.77 % 0.05 s 1 core @ 2.5 Ghz (Python)
190 YoloMono3D 78.50 % 91.43 % 58.80 % 0.05 s GPU @ 2.5 Ghz (Python)
191 3D-GCK 78.44 % 88.59 % 66.28 % 24 ms Tesla V100
N. Gählert, J. Wan, N. Jourdan, J. Finkbeiner, U. Franke and J. Denzler: Single-Shot 3D Detection of Vehicles from Monocular RGB Images via Geometrically Constrained Keypoints in Real-Time. 2020 IEEE Intelligent Vehicles Symposium (IV) 2020.
192 ITS-MDPL 78.24 % 92.05 % 70.73 % 0.16 s GPU @ 2.5 Ghz (Python)
193 3D-SSMFCNN code 77.82 % 77.84 % 68.67 % 0.1 s GPU @ 1.5 Ghz (C/C++)
L. Novak: Vehicle Detection and Pose Estimation for Autonomous Driving. 2017.
194 DA-3Ddet 77.73 % 89.01 % 61.48 % 0.4 s GPU @ 2.5 Ghz (Python)
195 3DVP code 75.71 % 84.44 % 64.41 % 40 s 8 cores @ 3.5 Ghz (Matlab + C/C++)
Y. Xiang, W. Choi, Y. Lin and S. Savarese: Data-Driven 3D Voxel Patterns for Object Category Recognition. IEEE Conference on Computer Vision and Pattern Recognition 2015.
196 GS3D 75.63 % 85.79 % 61.85 % 2 s 1 core @ 2.5 Ghz (C/C++)
B. Li, W. Ouyang, L. Sheng, X. Zeng and X. Wang: GS3D: An Efficient 3D Object Detection Framework for Autonomous Driving. IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2019.
197 Pose-RCNN 75.41 % 89.49 % 63.57 % 2 s >8 cores @ 2.5 Ghz (Python)
M. Braun, Q. Rao, Y. Wang and F. Flohr: Pose-RCNN: Joint object detection and pose estimation using 3D object proposals. Intelligent Transportation Systems (ITSC), 2016 IEEE 19th International Conference on 2016.
198 SubCat code 75.26 % 83.31 % 59.55 % 0.7 s 6 cores @ 3.5 Ghz (Matlab + C/C++)
E. Ohn-Bar and M. Trivedi: Learning to Detect Vehicles by Clustering Appearance Patterns. T-ITS 2015.
199 3D FCN
This method makes use of Velodyne laser scans.
 74.54 % 86.65 % 67.73 % >5 s 1 core @ 2.5 Ghz (C/C++)
B. Li: 3D Fully Convolutional Network for Vehicle Detection in Point Cloud. IROS 2017.
200 OC Stereo
This method uses stereo information.
 code 73.34 % 86.86 % 61.37 % 0.35 s 1 core @ 2.5 Ghz (Python + C/C++)
A. Pon, J. Ku, C. Li and S. Waslander: Object-Centric Stereo Matching for 3D Object Detection. ICRA 2020.
201 RTS3D 72.74 % 80.36 % 63.65 % 0.03 s GPU @ 2.5 Ghz (Python)
202 ROI-10D 68.14 % 75.32 % 58.98 % 0.2 s GPU @ 3.5 Ghz (Python)
F. Manhardt, W. Kehl and A. Gaidon: ROI-10D: Monocular Lifting of 2D Detection to 6D Pose and Metric Shape. Computer Vision and Pattern Recognition (CVPR) 2019.
203 BirdNet+
This method makes use of Velodyne laser scans.
 code 67.65 % 91.82 % 65.11 % 0.1 s Titan Xp (PyTorch)
A. Barrera, C. Guindel, J. Beltrán and F. García: BirdNet+: End-to-End 3D Object Detection in LiDAR Bird's Eye View. arXiv:2003.04188 [cs.CV] 2020.
204 multi-task CNN 67.51 % 79.00 % 58.80 % 25.1 ms GPU @ 2.0 Ghz (Python)
M. Oeljeklaus, F. Hoffmann and T. Bertram: A Fast Multi-Task CNN for Spatial Understanding of Traffic Scenes. IEEE Intelligent Transportation Systems Conference 2018.
205 I2BEV 67.31 % 78.28 % 59.52 % 0.63 s GPU @ 2.5 Ghz (Python)
206 Decoupled-3D 67.23 % 87.34 % 53.84 % 0.08 s GPU @ 2.5 Ghz (C/C++)
Y. Cai, B. Li, Z. Jiao, H. Li, X. Zeng and X. Wang: Monocular 3D Object Detection with Decoupled Structured Polygon Estimation and Height-Guided Depth Estimation. AAAI 2020.
207 BdCost48LDCF code 65.50 % 80.44 % 51.24 % 0.5 s 8 cores @ 3.5 Ghz (Matlab + C/C++)
A. Fernández-Baldera, J. Buenaposada and L. Baumela: BAdaCost: Multi-class Boosting with Costs . Pattern Recognition 2018.
208 OC-DPM 65.32 % 77.35 % 51.00 % 10 s 8 cores @ 2.5 Ghz (Matlab)
B. Pepik, M. Stark, P. Gehler and B. Schiele: Occlusion Patterns for Object Class Detection. IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2013.
209 deprecated 65.30 % 69.02 % 63.66 % 0.05 s GPU @ >3.5 Ghz (Python)
210 RefinedMPL 64.02 % 87.95 % 52.06 % 0.15 s GPU @ 2.5 Ghz (Python + C/C++)
J. Vianney, S. Aich and B. Liu: RefinedMPL: Refined Monocular PseudoLiDAR for 3D Object Detection in Autonomous Driving. arXiv preprint arXiv:1911.09712 2019.
211 DPM-VOC+VP 63.58 % 79.09 % 46.59 % 8 s 1 core @ 2.5 Ghz (C/C++)
B. Pepik, M. Stark, P. Gehler and B. Schiele: Multi-view and 3D Deformable Part Models. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) 2015.
212 AOG-View 62.62 % 77.62 % 48.27 % 3 s 1 core @ 2.5 Ghz (Matlab, C/C++)
B. Li, T. Wu and S. Zhu: Integrating Context and Occlusion for Car Detection by Hierarchical And-Or Model. ECCV 2014.
213 PG-MonoNet 61.20 % 70.34 % 52.59 % 0.19 s GPU @ 2.5 Ghz (Python)
214 LSVM-MDPM-sv 57.48 % 70.23 % 42.54 % 10 s 4 cores @ 3.0 Ghz (C/C++)
P. Felzenszwalb, R. Girshick, D. McAllester and D. Ramanan: Object Detection with Discriminatively Trained Part-Based Models. PAMI 2010.
A. Geiger, C. Wojek and R. Urtasun: Joint 3D Estimation of Objects and Scene Layout. NIPS 2011.
215 SAMME48LDCF code 57.26 % 76.28 % 43.55 % 0.5 s 8 cores @ 3.5 Ghz (Matlab + C/C++)
A. Fernández-Baldera, J. Buenaposada and L. Baumela: BAdaCost: Multi-class Boosting with Costs . Pattern Recognition 2018.
216 deprecated 57.01 % 62.54 % 54.94 % - -
217 BirdNet
This method makes use of Velodyne laser scans.
 56.94 % 79.20 % 54.88 % 0.11 s Titan Xp (Caffe)
J. Beltrán, C. Guindel, F. Moreno, D. Cruzado, F. García and A. Escalera: BirdNet: A 3D Object Detection Framework from LiDAR Information. 2018 21st International Conference on Intelligent Transportation Systems (ITSC) 2018.
218 DEFT 51.66 % 57.41 % 50.02 % 1 s GPU @ 2.5 Ghz (Python)
219 VeloFCN
This method makes use of Velodyne laser scans.
 51.05 % 70.03 % 44.82 % 1 s GPU @ 2.5 Ghz (Python + C/C++)
B. Li, T. Zhang and T. Xia: Vehicle Detection from 3D Lidar Using Fully Convolutional Network. RSS 2016 .
220 LZnet 50.95 % 52.89 % 49.73 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
221 Mono3D_PLiDAR code 49.39 % 76.90 % 41.13 % 0.1 s NVIDIA GeForce 1080 (pytorch)
X. Weng and K. Kitani: Monocular 3D Object Detection with Pseudo-LiDAR Point Cloud. arXiv:1903.09847 2019.
222 DPM-C8B1
This method uses stereo information.
 48.00 % 57.76 % 35.52 % 15 s 4 cores @ 2.5 Ghz (Matlab + C/C++)
J. Yebes, L. Bergasa and M. García-Garrido: Visual Object Recognition with 3D-Aware Features in KITTI Urban Scenes. Sensors 2015.
J. Yebes, L. Bergasa, R. Arroyo and A. Lázaro: Supervised learning and evaluation of KITTI's cars detector with DPM. IV 2014.
223 LTN 46.54 % 48.96 % 41.58 % 0.4 s GPU @ >3.5 Ghz (Python)
T. Wang, X. He, Y. Cai and G. Xiao: Learning a Layout Transfer Network for Context Aware Object Detection. IEEE Transactions on Intelligent Transportation Systems 2019.
224 sensekitti code 46.12 % 49.16 % 42.79 % 4.5 s GPU @ 2.5 Ghz (Python + C/C++)
B. Yang, J. Yan, Z. Lei and S. Li: Craft Objects from Images. CVPR 2016.
225 Kinematic3D code 45.50 % 58.33 % 34.81 % 0.12 s 1 core @ 1.5 Ghz (C/C++)
G. Brazil, G. Pons-Moll, X. Liu and B. Schiele: Kinematic 3D Object Detection in Monocular Video. ECCV 2020 .
226 PVGNet 40.79 % 43.04 % 39.42 % 0.05 s 1 core @ >3.5 Ghz (C/C++)
227 anonymous 40.75 % 45.00 % 34.48 % 1 s 1 core @ 2.5 Ghz (C/C++)
228 Dccnet 40.44 % 37.79 % 38.54 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
229 Chovy 40.34 % 41.64 % 38.31 % 0.04 s GPU @ 2.5 Ghz (Python)
230 cvMax 40.31 % 41.97 % 37.57 % 0.04 s GPU @ >3.5 Ghz (Python)
231 deprecated 40.03 % 40.31 % 37.35 % 0.04 s GPU @ 2.5 Ghz (Python)
232 3D-CVF at SPA
This method makes use of Velodyne laser scans.
 39.79 % 40.44 % 36.10 % 0.06 s 1 core @ 2.5 Ghz (C/C++)
J. Yoo, Y. Kim, J. Kim and J. Choi: 3D-CVF: Generating Joint Camera and LiDAR Features Using Cross-View Spatial Feature Fusion for 3D Object Detection. ECCV 2020.
233 CDI3D 39.62 % 41.27 % 34.88 % 0.03 s GPU @ 2.5 Ghz (Python)
234 HR-faster-rcnn 39.35 % 39.78 % 36.47 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
235 deprecated 38.89 % 40.49 % 35.13 % 0.06 s GPU @ >3.5 Ghz (Python)
236 dgist_multiDetNet 38.76 % 39.75 % 35.38 % 0.08 s GPU Titanx Pascal (Python)
237 BLPNet_V2 38.66 % 39.39 % 38.36 % 0.04 s 1 core @ 2.5 Ghz (Python)
238 PVF-NET 38.53 % 39.57 % 38.23 % 0.1 s 1 core @ 2.5 Ghz (Python)
239 SA-SSD code 38.30 % 39.40 % 37.07 % 0.04 s 1 core @ 2.5 Ghz (Python)
C. He, H. Zeng, J. Huang, X. Hua and L. Zhang: Structure Aware Single-stage 3D Object Detection from Point Cloud. CVPR 2020.
240 HR-Cascade-RCNN 38.25 % 39.72 % 35.92 % 0.3 s 1 core @ 2.5 Ghz (C/C++)
241 Faster RCNN + A 37.92 % 39.50 % 33.85 % 0.19 s GPU @ 2.5 Ghz (Python + C/C++)
242 KNN-GCNN 37.80 % 38.80 % 36.52 % 0.4 s 1 core @ 2.5 Ghz (Python + C/C++)
243 JSU-NET 37.60 % 41.33 % 33.41 % 0.1 s 1 core @ 2.5 Ghz (Python)
244 Faster RCNN + G 37.49 % 39.05 % 33.40 % 1.1 s GPU @ 2.5 Ghz (Python)
245 Faster RCNN + A 37.35 % 38.75 % 33.38 % 0.19 s GPU @ 2.5 Ghz (Python)
246 yolo4 37.27 % 38.19 % 32.45 % 0.02 s 1 core @ 2.5 Ghz (Python)
247 Point-GNN
This method makes use of Velodyne laser scans.
 code 37.20 % 38.66 % 36.29 % 0.6 s GPU @ 2.5 Ghz (Python)
W. Shi and R. Rajkumar: Point-GNN: Graph Neural Network for 3D Object Detection in a Point Cloud. CVPR 2020.
248 PP-3D 37.20 % 38.66 % 36.29 % 0.1 s 1 core @ 2.5 Ghz (Python)
249 F-3DNet 37.18 % 38.58 % 36.44 % 0.5 s GPU @ 2.5 Ghz (Python)
250 yolo4_5l 37.14 % 37.92 % 32.31 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
251 GAFM 37.08 % 40.28 % 33.08 % 0.5 s 1 core @ 2.5 Ghz (Python)
252 CRCNNA 37.04 % 40.19 % 32.03 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
253 Faster RCNN + Gr + A 36.95 % 38.22 % 33.16 % 1.29 s GPU @ 2.5 Ghz (Python)
254 yolo4_5l code 36.81 % 37.14 % 33.24 % 0.02 s 1 core @ 2.5 Ghz (Python + C/C++)
255 GA_BALANCE 36.62 % 38.44 % 31.94 % 1 s 1 core @ 2.5 Ghz (Python)
256 GA_rpn500 36.54 % 38.33 % 32.67 % 1 s 1 core @ 2.5 Ghz (Python)
257 GA2500 36.54 % 38.33 % 32.67 % 0.2 s 1 core @ 2.5 Ghz (Python)
258 GA_FULLDATA 36.43 % 38.90 % 31.61 % 1 s 4 cores @ 2.5 Ghz (Python)
259 PMN 36.23 % 38.08 % 32.05 % 0.2 s 1 core @ 2.5 Ghz (Python)
260 bigger_ga 36.21 % 38.41 % 31.58 % 1 s 1 core @ 2.5 Ghz (Python)
261 MMCOM 36.08 % 39.58 % 32.06 % 0.04 s 1 core @ 2.5 Ghz (Python)
262 Scan_YOLO 36.02 % 36.78 % 32.65 % 0.1 s 4 cores @ 3.0 Ghz (Python)
263 Multi-task DG 35.50 % 38.34 % 30.85 % 0.06 s GPU @ 2.5 Ghz (Python)
264 yolo_rgb 35.23 % 36.60 % 31.70 % 0.07 s GPU @ 2.5 Ghz (Python)
265 ga50 34.95 % 38.21 % 30.29 % 1 s 1 core @ 2.5 Ghz (Python)
266 bifpn_fsrn 33.84 % 37.56 % 29.98 % 0.07 s 1 core @ 2.5 Ghz (Python + C/C++)
267 Mag 33.74 % 38.34 % 28.76 % 0.07 s 1 core @ 2.5 Ghz (C/C++)
268 CBNet 32.63 % 36.51 % 29.26 % 1 s 4 cores @ 2.5 Ghz (Python)
269 MTNAS 31.15 % 35.43 % 27.02 % 0.02 s 1 core @ 2.5 Ghz (python)
270 EACV 31.15 % 35.61 % 27.63 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
271 yolo_depth 30.33 % 36.32 % 26.80 % 0.07 s GPU @ 2.5 Ghz (Python)
272 AOG code 29.81 % 33.28 % 23.91 % 3 s 4 cores @ 2.5 Ghz (Matlab)
T. Wu, B. Li and S. Zhu: Learning And-Or Models to Represent Context and Occlusion for Car Detection and Viewpoint Estimation. TPAMI 2016.
B. Li, T. Wu and S. Zhu: Integrating Context and Occlusion for Car Detection by Hierarchical And-Or Model. ECCV 2014.
273 m-prcnn
This method uses stereo information.
 29.62 % 34.80 % 22.79 % 0.43 s 1 core @ 2.5 Ghz (Python)
274 DAM 28.97 % 37.05 % 25.28 % 1 s GPU @ 2.5 Ghz (Python)
275 SubCat48LDCF code 26.68 % 34.33 % 19.44 % 0.5 s 8 cores @ 3.5 Ghz (Matlab + C/C++)
A. Fernández-Baldera, J. Buenaposada and L. Baumela: BAdaCost: Multi-class Boosting with Costs . Pattern Recognition 2018.
276 RT3D-GMP
This method uses stereo information.
 24.27 % 28.33 % 18.51 % 0.06 s GPU @ 2.5 Ghz (Python + C/C++)
277 RT3DStereo
This method uses stereo information.
 21.41 % 25.58 % 17.52 % 0.08 s GPU @ 2.5 Ghz (C/C++)
H. Königshof, N. Salscheider and C. Stiller: Realtime 3D Object Detection for Automated Driving Using Stereo Vision and Semantic Information. Proc. IEEE Intl. Conf. Intelligent Transportation Systems 2019.
278 CSoR
This method makes use of Velodyne laser scans.
 code 20.82 % 30.65 % 17.14 % 3.5 s 4 cores @ >3.5 Ghz (Python + C/C++)
L. Plotkin: PyDriver: Entwicklung eines Frameworks für räumliche Detektion und Klassifikation von Objekten in Fahrzeugumgebung. 2015.
279 RT3D
This method makes use of Velodyne laser scans.
 18.96 % 24.41 % 19.85 % 0.09 s GPU @ 1.8Ghz
Y. Zeng, Y. Hu, S. Liu, J. Ye, Y. Han, X. Li and N. Sun: RT3D: Real-Time 3-D Vehicle Detection in LiDAR Point Cloud for Autonomous Driving. IEEE Robotics and Automation Letters 2018.
280 VoxelJones code 15.41 % 17.83 % 14.13 % .18 s 1 core @ 2.5 Ghz (Python + C/C++)
M. Motro and J. Ghosh: Vehicular Multi-object Tracking with Persistent Detector Failures. arXiv preprint arXiv:1907.11306 2019.
281 MuRF 1.75 % 0.63 % 2.14 % 0.05 s GPU @ 1.5 Ghz (Python + C/C++)
282 PiP 1.45 % 0.56 % 1.85 % 0.033 s 1 core @ 2.5 Ghz (Python)
283 Simple3D Net 1.38 % 0.63 % 1.76 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
284 Associate-3Ddet code 1.20 % 0.52 % 1.38 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
L. Du, X. Ye, X. Tan, J. Feng, Z. Xu, E. Ding and S. Wen: Associate-3Ddet: Perceptual-to-Conceptual Association for 3D Point Cloud Object Detection. The IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2020.
Table as LaTeX | Only published Methods


Pedestrians


Method Setting Code Moderate Easy Hard Runtime Environment
1 VMVS
This method makes use of Velodyne laser scans.
 68.19 % 79.98 % 63.18 % 0.25 s GPU @ 2.5 Ghz (Python)
J. Ku, A. Pon, S. Walsh and S. Waslander: Improving 3D object detection for pedestrians with virtual multi-view synthesis orientation estimation. IROS 2019.
2 SubCNN 66.70 % 79.65 % 61.35 % 2 s GPU @ 3.5 Ghz (Python + C/C++)
Y. Xiang, W. Choi, Y. Lin and S. Savarese: Subcategory-aware Convolutional Neural Networks for Object Proposals and Detection. IEEE Winter Conference on Applications of Computer Vision (WACV) 2017.
3 F-ConvNet
This method makes use of Velodyne laser scans.
 code 63.87 % 75.19 % 58.57 % 0.47 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Wang and K. Jia: Frustum ConvNet: Sliding Frustums to Aggregate Local Point-Wise Features for Amodal 3D Object Detection. IROS 2019.
4 3DOP
This method uses stereo information.
 code 61.48 % 74.22 % 55.89 % 3s GPU @ 2.5 Ghz (Matlab + C/C++)
X. Chen, K. Kundu, Y. Zhu, A. Berneshawi, H. Ma, S. Fidler and R. Urtasun: 3D Object Proposals for Accurate Object Class Detection. NIPS 2015.
5 HotSpotNet 60.65 % 70.36 % 57.42 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
Q. Chen, L. Sun, Z. Wang, K. Jia and A. Yuille: object as hotspots. Proceedings of the European Conference on Computer Vision (ECCV) 2020.
6 DeepStereoOP 60.15 % 73.76 % 55.30 % 3.4 s GPU @ 3.5 Ghz (Matlab + C/C++)
C. Pham and J. Jeon: Robust Object Proposals Re-ranking for Object Detection in Autonomous Driving Using Convolutional Neural Networks. Signal Processing: Image Communiation 2017.
7 Pose-RCNN 59.84 % 76.24 % 53.59 % 2 s >8 cores @ 2.5 Ghz (Python)
M. Braun, Q. Rao, Y. Wang and F. Flohr: Pose-RCNN: Joint object detection and pose estimation using 3D object proposals. Intelligent Transportation Systems (ITSC), 2016 IEEE 19th International Conference on 2016.
8 Mono3CN 59.17 % 72.16 % 53.22 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
9 FFNet code 58.87 % 69.24 % 53.75 % 1.07 s GPU @ 1.5 Ghz (Python)
C. Zhao, Y. Qian and M. Yang: Monocular Pedestrian Orientation Estimation Based on Deep 2D-3D Feedforward. Pattern Recognition 2019.
10 Mono3D code 58.66 % 71.19 % 53.94 % 4.2 s GPU @ 2.5 Ghz (Matlab + C/C++)
X. Chen, K. Kundu, Z. Zhang, H. Ma, S. Fidler and R. Urtasun: Monocular 3D Object Detection for Autonomous Driving. CVPR 2016.
11 MVX-Net++ 54.86 % 64.23 % 50.85 % 0.15 s 1 core @ 2.5 Ghz (C/C++)
12 RethinkDet3D 54.72 % 63.97 % 50.72 % 0.15 s 1 core @ 2.5 Ghz (Python)
13 MonoPSR code 54.65 % 68.98 % 50.07 % 0.2 s GPU @ 3.5 Ghz (Python)
J. Ku*, A. Pon* and S. Waslander: Monocular 3D Object Detection Leveraging Accurate Proposals and Shape Reconstruction. CVPR 2019.
14 Deformable PV-RCNN
This method makes use of Velodyne laser scans.
 code 54.38 % 63.12 % 51.98 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
P. Bhattacharyya and K. Czarnecki: Deformable PV-RCNN: Improving 3D Object Detection with Learned Deformations. ECCV 2020 Perception for Autonomous Driving Workshop.
15 CA-RCNN 53.53 % 63.48 % 50.30 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
16 AP-RCNN 53.38 % 65.13 % 49.49 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
17 SVGA-Net
This method makes use of Velodyne laser scans.
 53.36 % 63.39 % 50.43 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
18 Baseline of CA RCNN 53.35 % 63.39 % 50.42 % 0.1 s GPU @ 2.5 Ghz (Python)
19 CVIS-DF3D 53.35 % 63.39 % 50.42 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
20 IC-SECOND 52.63 % 61.60 % 50.07 % 0.06 s 1 core @ 2.5 Ghz (C/C++)
21 MMLab PV-RCNN
This method makes use of Velodyne laser scans.
 code 52.42 % 63.45 % 49.23 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
S. Shi, C. Guo, L. Jiang, Z. Wang, J. Shi, X. Wang and H. Li: PV-RCNN: Point-Voxel Feature Set Abstraction for 3D Object Detection. CVPR 2020.
22 IC-PVRCNN 52.40 % 61.31 % 50.15 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
23 MMLab-PartA^2
This method makes use of Velodyne laser scans.
 code 52.20 % 63.51 % 48.27 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
S. Shi, Z. Wang, J. Shi, X. Wang and H. Li: From Points to Parts: 3D Object Detection from Point Cloud with Part-aware and Part-aggregation Network. IEEE Transactions on Pattern Analysis and Machine Intelligence 2020.
24 FRCNN+Or code 52.15 % 67.03 % 47.14 % 0.09 s Titan Xp GPU
C. Guindel, D. Martin and J. Armingol: Fast Joint Object Detection and Viewpoint Estimation for Traffic Scene Understanding. IEEE Intelligent Transportation Systems Magazine 2018.
C. Guindel, D. Martin and J. Armingol: Joint Object Detection and Viewpoint Estimation using CNN features. IEEE International Conference on Vehicular Electronics and Safety (ICVES) 2017.
25 MGACNet 50.52 % 60.32 % 47.92 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
26 CVIS-DF3D_v2 50.51 % 60.74 % 47.69 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
27 PointPainting
This method makes use of Velodyne laser scans.
 50.22 % 59.25 % 46.95 % 0.4 s GPU @ 2.5 Ghz (Python + C/C++)
S. Vora, A. Lang, B. Helou and O. Beijbom: PointPainting: Sequential Fusion for 3D Object Detection. CVPR 2020.
28 deprecated 48.72 % 57.13 % 46.45 % 0.06 s 1 core @ 2.5 Ghz (C/C++)
29 ARPNET 48.49 % 60.47 % 45.02 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
Y. Ye, C. Zhang and X. Hao: ARPNET: attention region proposal network for 3D object detection. Science China Information Sciences 2019.
30 3DBN_2 48.43 % 59.19 % 45.73 % 0.12 s 1 core @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
31 TBD 48.34 % 58.57 % 44.85 % 0.05 s GPU @ 2.5 Ghz (Python)
32 PointPillars
This method makes use of Velodyne laser scans.
 code 48.05 % 57.47 % 45.40 % 16 ms 1080ti GPU and Intel i7 CPU
A. Lang, S. Vora, H. Caesar, L. Zhou, J. Yang and O. Beijbom: PointPillars: Fast Encoders for Object Detection from Point Clouds. CVPR 2019.
33 PPFNet code 47.73 % 55.78 % 44.56 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
34 MMLab-PointRCNN
This method makes use of Velodyne laser scans.
 code 47.33 % 57.19 % 44.31 % 0.1 s GPU @ 2.5 Ghz (Python + C/C++)
S. Shi, X. Wang and H. Li: Pointrcnn: 3d object proposal generation and detection from point cloud. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2019.
35 PVNet 46.68 % 57.18 % 44.38 % 0,1 s 1 core @ 2.5 Ghz (Python)
36 Shift R-CNN (mono) code 46.56 % 64.73 % 41.86 % 0.25 s GPU @ 1.5 Ghz (Python)
A. Naiden, V. Paunescu, G. Kim, B. Jeon and M. Leordeanu: Shift R-CNN: Deep Monocular 3D Object Detection With Closed-form Geometric Constraints. ICIP 2019.
37 IGRP+ 46.34 % 57.61 % 42.75 % 0.18 s 1 core @ 2.5 Ghz (Python + C/C++)
38 VOXEL_FPN_HR 45.65 % 56.17 % 42.10 % 0.12 s 8 cores @ 2.5 Ghz (Python)
ERROR: Wrong syntax in BIBTEX file.
39 UDI-mono3D 45.37 % 57.89 % 40.78 % 0.05 s 1 core @ 2.5 Ghz (Python)
40 UDI-mono3D 44.75 % 57.42 % 40.21 % 0.05 s 1 core @ 2.5 Ghz (Python)
41 HWFD 44.66 % 48.89 % 42.14 % 0.21 s one 1080Ti
42 SS3D_HW 44.43 % 59.56 % 38.77 % 0.4 s GPU @ 2.5 Ghz (Python)
43 SRDL
This method uses stereo information.
This method makes use of Velodyne laser scans.
 44.31 % 53.68 % 42.32 % 0.15 s GPU @ 2.5 Ghz (Python + C/C++)
44 AVOD-FPN
This method makes use of Velodyne laser scans.
 code 43.99 % 53.48 % 41.56 % 0.1 s Titan X (Pascal)
J. Ku, M. Mozifian, J. Lee, A. Harakeh and S. Waslander: Joint 3D Proposal Generation and Object Detection from View Aggregation. IROS 2018.
45 Disp R-CNN (velo)
This method uses stereo information.
 code 43.99 % 60.06 % 39.79 % 0.42 s GPU @ 2.5 Ghz (Python + C/C++)
J. Sun, L. Chen, Y. Xie, S. Zhang, Q. Jiang, X. Zhou and H. Bao: Disp R-CNN: Stereo 3D Object Detection via Shape Prior Guided Instance Disparity Estimation. CVPR 2020.
46 Disp R-CNN
This method uses stereo information.
 code 43.76 % 60.00 % 39.55 % 0.42 s GPU @ 2.5 Ghz (Python + C/C++)
J. Sun, L. Chen, Y. Xie, S. Zhang, Q. Jiang, X. Zhou and H. Bao: Disp R-CNN: Stereo 3D Object Detection via Shape Prior Guided Instance Disparity Estimation. CVPR 2020.
47 dgist_multiDetNet 43.48 % 49.02 % 40.97 % 0.08 s GPU Titanx Pascal (Python)
48 PiP 42.76 % 51.23 % 40.06 % 0.033 s 1 core @ 2.5 Ghz (Python)
49 MonoPair 42.38 % 55.26 % 38.53 % 0.06 s GPU @ 2.5 Ghz (Python + C/C++)
Y. Chen, L. Tai, K. Sun and M. Li: MonoPair: Monocular 3D Object Detection Using Pairwise Spatial Relationships. The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2020.
50 VICNet 42.35 % 49.90 % 39.38 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
51 NLK-3D 41.71 % 54.22 % 39.32 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
52 MTMono3d 41.63 % 54.28 % 36.32 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
53 PFF3D
This method makes use of Velodyne laser scans.
 40.99 % 48.75 % 38.99 % 0.05 s GPU @ 3.0 Ghz (Python + C/C++)
54 Faster RCNN + Gr + A 40.92 % 47.81 % 37.89 % 1.29 s GPU @ 2.5 Ghz (Python)
55 HR-SECOND code 40.81 % 51.12 % 37.48 % 0.11 s 1 core @ 2.5 Ghz (Python + C/C++)
56 Faster RCNN + G 40.49 % 47.16 % 37.57 % 1.1 s GPU @ 2.5 Ghz (Python)
57 Faster RCNN + A 39.95 % 47.52 % 37.08 % 0.19 s GPU @ 2.5 Ghz (Python)
58 CentrNet-FG 39.88 % 47.51 % 37.49 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
59 AB3DMOT
This method makes use of Velodyne laser scans.
This is an online method (no batch processing).
 code 39.76 % 50.30 % 36.90 % 0.0047s 1 core @ 2.5 Ghz (python)
X. Weng and K. Kitani: A Baseline for 3D Multi-Object Tracking. arXiv:1907.03961 2019.
60 FCY
This method makes use of Velodyne laser scans.
 39.67 % 51.30 % 35.90 % 0.02 s GPU @ 2.5 Ghz (Python)
61 SS3D 39.60 % 53.72 % 35.40 % 48 ms Tesla V100 (Python)
E. Jörgensen, C. Zach and F. Kahl: Monocular 3D Object Detection and Box Fitting Trained End-to-End Using Intersection-over-Union Loss. CoRR 2019.
62 Faster RCNN + A 39.44 % 46.80 % 36.46 % 0.19 s GPU @ 2.5 Ghz (Python + C/C++)
63 NLK-ALL code 39.31 % 49.20 % 35.60 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
64 MMCOM 39.20 % 46.12 % 36.81 % 0.04 s 1 core @ 2.5 Ghz (Python)
65 HR-faster-rcnn 39.02 % 47.41 % 35.57 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
66 SemanticVoxels 38.95 % 45.59 % 37.21 % 0.04 s GPU @ 2.5 Ghz (Python + C/C++)
J. Fei, W. Chen, P. Heidenreich, S. Wirges and C. Stiller: SemanticVoxels: Sequential Fusion for 3D Pedestrian Detection using LiDAR Point Cloud and Semantic Segmentation. MFI 2020.
67 Multi-task DG 38.79 % 46.21 % 36.07 % 0.06 s GPU @ 2.5 Ghz (Python)
68 Center3D 38.59 % 53.15 % 34.77 % 0.05 s GPU @ 3.5 Ghz (Python)
69 DAMNET code 37.88 % 49.72 % 35.52 % 1 s 1 core @ 2.5 Ghz (C/C++)
70 DPM-VOC+VP 37.79 % 52.91 % 33.27 % 8 s 1 core @ 2.5 Ghz (C/C++)
B. Pepik, M. Stark, P. Gehler and B. Schiele: Multi-view and 3D Deformable Part Models. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) 2015.
71 CG-Stereo
This method uses stereo information.
 36.47 % 48.23 % 32.77 % 0.57 s GeForce RTX 2080 Ti
C. Li, J. Ku and S. Waslander: Confidence Guided Stereo 3D Object Detection with Split Depth Estimation. IROS 2020.
72 TANet code 36.21 % 42.54 % 34.39 % 0.035s GPU @ 2.5 Ghz (Python + C/C++)
Z. Liu, X. Zhao, T. Huang, R. Hu, Y. Zhou and X. Bai: TANet: Robust 3D Object Detection from Point Clouds with Triple Attention. AAAI 2020.
73 Stereo3D
This method uses stereo information.
 35.62 % 48.99 % 31.58 % 0.1 s GPU 1080Ti
74 PMN 35.57 % 44.42 % 32.68 % 0.2 s 1 core @ 2.5 Ghz (Python)
75 SCNet
This method makes use of Velodyne laser scans.
 35.49 % 44.50 % 33.38 % 0.04 s GPU @ 3.0 Ghz (Python)
Z. Wang, H. Fu, L. Wang, L. Xiao and B. Dai: SCNet: Subdivision Coding Network for Object Detection Based on 3D Point Cloud. IEEE Access 2019.
76 NL_M3D 35.20 % 46.64 % 30.56 % 0.2 s 1 core @ 2.5 Ghz (Python)
77 CRCNNA 34.88 % 43.18 % 31.90 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
78 sensekitti code 34.26 % 41.03 % 31.51 % 4.5 s GPU @ 2.5 Ghz (Python + C/C++)
B. Yang, J. Yan, Z. Lei and S. Li: Craft Objects from Images. CVPR 2016.
79 Pointpillar_TV 34.24 % 42.95 % 32.08 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
80 vxnet 33.81 % 47.20 % 30.15 % 1 s 1 core @ 2.5 Ghz (C/C++)
81 D4LCN code 33.62 % 46.73 % 28.71 % 0.2 s GPU @ 2.5 Ghz (Python + C/C++)
M. Ding, Y. Huo, H. Yi, Z. Wang, J. Shi, Z. Lu and P. Luo: Learning Depth-Guided Convolutions for Monocular 3D Object Detection. CVPR 2020.
82 JSU-NET 33.55 % 45.79 % 30.72 % 0.1 s 1 core @ 2.5 Ghz (Python)
83 DP3D 33.35 % 46.50 % 29.89 % 0.05 s GPU @ 1.5 Ghz (Python + C/C++)
84 SparsePool code 33.35 % 43.86 % 29.99 % 0.13 s 8 cores @ 2.5 Ghz (Python)
Z. Wang, W. Zhan and M. Tomizuka: Fusing bird view lidar point cloud and front view camera image for deep object detection. arXiv preprint arXiv:1711.06703 2017.
85 SparsePool code 33.29 % 43.52 % 30.01 % 0.13 s 8 cores @ 2.5 Ghz (Python)
Z. Wang, W. Zhan and M. Tomizuka: Fusing bird view lidar point cloud and front view camera image for deep object detection. arXiv preprint arXiv:1711.06703 2017.
86 LSVM-MDPM-sv 33.01 % 45.60 % 29.27 % 10 s 4 cores @ 3.0 Ghz (C/C++)
P. Felzenszwalb, R. Girshick, D. McAllester and D. Ramanan: Object Detection with Discriminatively Trained Part-Based Models. PAMI 2010.
A. Geiger, C. Wojek and R. Urtasun: Joint 3D Estimation of Objects and Scene Layout. NIPS 2011.
87 DP3D 32.99 % 44.19 % 28.19 % 0.07 s GPU @ 1.5 Ghz (Python + C/C++)
88 AVOD
This method makes use of Velodyne laser scans.
 code 32.19 % 42.54 % 29.09 % 0.08 s Titan X (pascal)
J. Ku, M. Mozifian, J. Lee, A. Harakeh and S. Waslander: Joint 3D Proposal Generation and Object Detection from View Aggregation. IROS 2018.
89 Complexer-YOLO
This method makes use of Velodyne laser scans.
 32.13 % 37.32 % 28.94 % 0.06 s GPU @ 3.5 Ghz (C/C++)
M. Simon, K. Amende, A. Kraus, J. Honer, T. Samann, H. Kaulbersch, S. Milz and H. Michael Gross: Complexer-YOLO: Real-Time 3D Object Detection and Tracking on Semantic Point Clouds. The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops 2019.
90 RPN+BF code 32.12 % 41.19 % 28.83 % 0.6 s GPU @ 2.5 Ghz (Matlab + C/C++)
L. Zhang, L. Lin, X. Liang and K. He: Is Faster R-CNN Doing Well for Pedestrian Detection?. ECCV 2016.
91 MP-Mono 32.02 % 44.19 % 28.72 % 0.16 s GPU @ 2.5 Ghz (Python)
92 KNN-GCNN 31.91 % 39.25 % 29.76 % 0.4 s 1 core @ 2.5 Ghz (Python + C/C++)
93 M3D-RPN code 31.88 % 44.33 % 28.55 % 0.16 s GPU @ 1.5 Ghz (Python)
G. Brazil and X. Liu: M3D-RPN: Monocular 3D Region Proposal Network for Object Detection . ICCV 2019 .
94 Point-GNN
This method makes use of Velodyne laser scans.
 code 31.86 % 39.16 % 29.65 % 0.6 s GPU @ 2.5 Ghz (Python)
W. Shi and R. Rajkumar: Point-GNN: Graph Neural Network for 3D Object Detection in a Point Cloud. CVPR 2020.
95 PP-3D 31.86 % 39.16 % 29.65 % 0.1 s 1 core @ 2.5 Ghz (Python)
96 yolo4_5l 31.53 % 40.97 % 27.63 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
97 SubCat 31.26 % 42.31 % 27.39 % 1.2 s 6 cores @ 2.5 Ghz (Matlab + C/C++)
E. Ohn-Bar and M. Trivedi: Fast and Robust Object Detection Using Visual Subcategories. Computer Vision and Pattern Recognition Workshops Mobile Vision 2014.
98 DAM 30.58 % 41.32 % 27.84 % 1 s GPU @ 2.5 Ghz (Python)
99 Mag 30.28 % 39.29 % 27.59 % 0.07 s 1 core @ 2.5 Ghz (C/C++)
100 yolo4 30.09 % 40.84 % 27.35 % 0.02 s 1 core @ 2.5 Ghz (Python)
101 PB3D
This method uses stereo information.
 30.02 % 40.27 % 26.86 % 0.42 s 1 core @ 2.5 Ghz (C/C++)
102 PG-MonoNet 29.56 % 37.28 % 26.48 % 0.19 s GPU @ 2.5 Ghz (Python)
103 BirdNet+
This method makes use of Velodyne laser scans.
 code 29.56 % 36.76 % 28.10 % 0.1 s Titan Xp (PyTorch)
A. Barrera, C. Guindel, J. Beltrán and F. García: BirdNet+: End-to-End 3D Object Detection in LiDAR Bird's Eye View. arXiv:2003.04188 [cs.CV] 2020.
104 CDI3D 29.51 % 36.82 % 27.23 % 0.03 s GPU @ 2.5 Ghz (Python)
105 LZnet 28.91 % 35.42 % 26.86 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
106 yolo4_5l code 28.60 % 38.95 % 25.97 % 0.02 s 1 core @ 2.5 Ghz (Python + C/C++)
107 yolo_depth 28.06 % 38.75 % 25.37 % 0.07 s GPU @ 2.5 Ghz (Python)
108 LAPNet 27.11 % 37.41 % 24.04 % 0.03 s 1 core @ 2.5 Ghz (Python)
109 yolo_rgb 26.85 % 35.91 % 24.37 % 0.07 s GPU @ 2.5 Ghz (Python)
110 EACV 25.81 % 34.71 % 23.36 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
111 DSGN
This method uses stereo information.
 code 24.32 % 31.21 % 23.09 % 0.67 s NVIDIA Tesla V100
Y. Chen, S. Liu, X. Shen and J. Jia: DSGN: Deep Stereo Geometry Network for 3D Object Detection. CVPR 2020.
112 ACF 24.31 % 32.23 % 21.70 % 1 s 1 core @ 3.5 Ghz (Matlab + C/C++)
P. Doll\'ar, R. Appel, S. Belongie and P. Perona: Fast Feature Pyramids for Object Detection. PAMI 2014.
113 multi-task CNN 22.80 % 30.30 % 20.47 % 25.1 ms GPU @ 2.0 Ghz (Python)
M. Oeljeklaus, F. Hoffmann and T. Bertram: A Fast Multi-Task CNN for Spatial Understanding of Traffic Scenes. IEEE Intelligent Transportation Systems Conference 2018.
114 ACF-MR 22.61 % 29.23 % 20.08 % 0.6 s 1 core @ 3.5 Ghz (C/C++)
R. Rajaram, E. Ohn-Bar and M. Trivedi: Looking at Pedestrians at Different Scales: A Multi-resolution Approach and Evaluations. T-ITS 2016.
115 OC Stereo
This method uses stereo information.
 code 22.02 % 31.36 % 20.20 % 0.35 s 1 core @ 2.5 Ghz (Python + C/C++)
A. Pon, J. Ku, C. Li and S. Waslander: Object-Centric Stereo Matching for 3D Object Detection. ICRA 2020.
116 BirdNet
This method makes use of Velodyne laser scans.
 21.83 % 27.12 % 20.56 % 0.11 s Titan Xp (Caffe)
J. Beltrán, C. Guindel, F. Moreno, D. Cruzado, F. García and A. Escalera: BirdNet: A 3D Object Detection Framework from LiDAR Information. 2018 21st International Conference on Intelligent Transportation Systems (ITSC) 2018.
117 RT3D-GMP
This method uses stereo information.
 20.81 % 29.49 % 18.34 % 0.06 s GPU @ 2.5 Ghz (Python + C/C++)
118 DPM-C8B1
This method uses stereo information.
 19.17 % 27.79 % 16.48 % 15 s 4 cores @ 2.5 Ghz (Matlab + C/C++)
J. Yebes, L. Bergasa and M. García-Garrido: Visual Object Recognition with 3D-Aware Features in KITTI Urban Scenes. Sensors 2015.
J. Yebes, L. Bergasa, R. Arroyo and A. Lázaro: Supervised learning and evaluation of KITTI's cars detector with DPM. IV 2014.
119 RefinedMPL 17.26 % 25.83 % 15.41 % 0.15 s GPU @ 2.5 Ghz (Python + C/C++)
J. Vianney, S. Aich and B. Liu: RefinedMPL: Refined Monocular PseudoLiDAR for 3D Object Detection in Autonomous Driving. arXiv preprint arXiv:1911.09712 2019.
120 I2BEV 17.13 % 24.45 % 15.79 % 0.63 s GPU @ 2.5 Ghz (Python)
121 RT3DStereo
This method uses stereo information.
 15.34 % 21.41 % 13.23 % 0.08 s GPU @ 2.5 Ghz (C/C++)
H. Königshof, N. Salscheider and C. Stiller: Realtime 3D Object Detection for Automated Driving Using Stereo Vision and Semantic Information. Proc. IEEE Intl. Conf. Intelligent Transportation Systems 2019.
122 Simple3D Net 11.95 % 13.63 % 11.68 % 0.02 s GPU @ 2.5 Ghz (Python)
123 CBNet 0.72 % 0.56 % 0.75 % 1 s 4 cores @ 2.5 Ghz (Python)
124 UM3D_TUM 0.00 % 0.00 % 0.00 % 0.05 s 1 core @ 2.5 Ghz (Python)
Table as LaTeX | Only published Methods


Cyclists


Method Setting Code Moderate Easy Hard Runtime Environment
1 HRI-MSP-L
This method makes use of Velodyne laser scans.
 82.89 % 91.97 % 75.38 % 0.07 s 1 core @ 2.5 Ghz (C/C++)
2 Deformable PV-RCNN
This method makes use of Velodyne laser scans.
 code 80.05 % 88.52 % 74.20 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
P. Bhattacharyya and K. Czarnecki: Deformable PV-RCNN: Improving 3D Object Detection with Learned Deformations. ECCV 2020 Perception for Autonomous Driving Workshop.
3 MMLab PV-RCNN
This method makes use of Velodyne laser scans.
 code 79.70 % 86.43 % 72.96 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
S. Shi, C. Guo, L. Jiang, Z. Wang, J. Shi, X. Wang and H. Li: PV-RCNN: Point-Voxel Feature Set Abstraction for 3D Object Detection. CVPR 2020.
4 HotSpotNet 78.31 % 85.79 % 71.24 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
Q. Chen, L. Sun, Z. Wang, K. Jia and A. Yuille: object as hotspots. Proceedings of the European Conference on Computer Vision (ECCV) 2020.
5 IC-PVRCNN 77.69 % 88.39 % 71.45 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
6 MMLab-PartA^2
This method makes use of Velodyne laser scans.
 code 77.52 % 88.70 % 70.41 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
S. Shi, Z. Wang, J. Shi, X. Wang and H. Li: From Points to Parts: 3D Object Detection from Point Cloud with Part-aware and Part-aggregation Network. IEEE Transactions on Pattern Analysis and Machine Intelligence 2020.
7 PointPainting
This method makes use of Velodyne laser scans.
 76.92 % 87.33 % 68.21 % 0.4 s GPU @ 2.5 Ghz (Python + C/C++)
S. Vora, A. Lang, B. Helou and O. Beijbom: PointPainting: Sequential Fusion for 3D Object Detection. CVPR 2020.
8 TBD 76.79 % 87.00 % 70.00 % 0.05 s GPU @ 2.5 Ghz (Python)
9 F-ConvNet
This method makes use of Velodyne laser scans.
 code 76.71 % 86.39 % 66.92 % 0.47 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Wang and K. Jia: Frustum ConvNet: Sliding Frustums to Aggregate Local Point-Wise Features for Amodal 3D Object Detection. IROS 2019.
10 CVIS-DF3D_v2 76.47 % 86.19 % 69.81 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
11 IC-SECOND 75.91 % 85.69 % 70.39 % 0.06 s 1 core @ 2.5 Ghz (C/C++)
12 deprecated 75.78 % 83.89 % 70.36 % 0.06 s 1 core @ 2.5 Ghz (C/C++)
13 VOXEL_FPN_HR 74.77 % 87.41 % 68.16 % 0.12 s 8 cores @ 2.5 Ghz (Python)
ERROR: Wrong syntax in BIBTEX file.
14 MVX-Net++ 74.65 % 86.53 % 67.43 % 0.15 s 1 core @ 2.5 Ghz (C/C++)
15 RethinkDet3D 74.33 % 88.54 % 65.20 % 0.15 s 1 core @ 2.5 Ghz (Python)
16 SRDL
This method uses stereo information.
This method makes use of Velodyne laser scans.
 74.24 % 87.85 % 67.84 % 0.15 s GPU @ 2.5 Ghz (Python + C/C++)
17 3DBN_2 73.69 % 87.96 % 66.91 % 0.12 s 1 core @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
18 MGACNet 73.43 % 85.32 % 66.87 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
19 NLK-ALL code 73.32 % 86.61 % 66.56 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
20 Baseline of CA RCNN 73.22 % 85.17 % 66.44 % 0.1 s GPU @ 2.5 Ghz (Python)
21 CVIS-DF3D 73.22 % 85.17 % 66.44 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
22 SVGA-Net
This method makes use of Velodyne laser scans.
 73.21 % 85.22 % 66.45 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
23 CA-RCNN 72.99 % 84.91 % 65.60 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
24 MMLab-PointRCNN
This method makes use of Velodyne laser scans.
 code 72.81 % 85.94 % 65.84 % 0.1 s GPU @ 2.5 Ghz (Python + C/C++)
S. Shi, X. Wang and H. Li: Pointrcnn: 3d object proposal generation and detection from point cloud. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2019.
25 PiP 71.10 % 82.83 % 64.88 % 0.033 s 1 core @ 2.5 Ghz (Python)
26 PVNet 70.50 % 83.44 % 64.47 % 0,1 s 1 core @ 2.5 Ghz (Python)
27 NLK-3D 70.10 % 85.69 % 63.27 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
28 AP-RCNN 69.74 % 86.02 % 62.90 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
29 FCY
This method makes use of Velodyne laser scans.
 69.71 % 82.71 % 63.29 % 0.02 s GPU @ 2.5 Ghz (Python)
30 HR-SECOND code 69.60 % 82.42 % 62.47 % 0.11 s 1 core @ 2.5 Ghz (Python + C/C++)
31 AB3DMOT
This method makes use of Velodyne laser scans.
This is an online method (no batch processing).
 code 69.54 % 82.18 % 62.98 % 0.0047s 1 core @ 2.5 Ghz (Python)
X. Weng and K. Kitani: A Baseline for 3D Multi-Object Tracking. arXiv:1907.03961 2019.
32 ARPNET 68.72 % 82.61 % 62.00 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
Y. Ye, C. Zhang and X. Hao: ARPNET: attention region proposal network for 3D object detection. Science China Information Sciences 2019.
33 PointPillars
This method makes use of Velodyne laser scans.
 code 68.55 % 83.79 % 61.71 % 16 ms 1080ti GPU and Intel i7 CPU
A. Lang, S. Vora, H. Caesar, L. Zhou, J. Yang and O. Beijbom: PointPillars: Fast Encoders for Object Detection from Point Clouds. CVPR 2019.
34 IGRP+ 67.25 % 83.67 % 60.84 % 0.18 s 1 core @ 2.5 Ghz (Python + C/C++)
35 VICNet 66.83 % 86.12 % 59.74 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
36 CentrNet-FG 66.68 % 82.23 % 59.21 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
37 TANet code 66.37 % 81.15 % 60.10 % 0.035s GPU @ 2.5 Ghz (Python + C/C++)
Z. Liu, X. Zhao, T. Huang, R. Hu, Y. Zhou and X. Bai: TANet: Robust 3D Object Detection from Point Clouds with Triple Attention. AAAI 2020.
38 Pointpillar_TV 65.12 % 78.88 % 58.73 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
39 PFF3D
This method makes use of Velodyne laser scans.
 64.06 % 78.02 % 58.06 % 0.05 s GPU @ 3.0 Ghz (Python + C/C++)
40 SubCNN 63.36 % 71.97 % 55.42 % 2 s GPU @ 3.5 Ghz (Python + C/C++)
Y. Xiang, W. Choi, Y. Lin and S. Savarese: Subcategory-aware Convolutional Neural Networks for Object Proposals and Detection. IEEE Winter Conference on Applications of Computer Vision (WACV) 2017.
41 Pose-RCNN 62.02 % 75.74 % 53.99 % 2 s >8 cores @ 2.5 Ghz (Python)
M. Braun, Q. Rao, Y. Wang and F. Flohr: Pose-RCNN: Joint object detection and pose estimation using 3D object proposals. Intelligent Transportation Systems (ITSC), 2016 IEEE 19th International Conference on 2016.
42 SCNet
This method makes use of Velodyne laser scans.
 61.11 % 77.77 % 54.82 % 0.04 s GPU @ 3.0 Ghz (Python)
Z. Wang, H. Fu, L. Wang, L. Xiao and B. Dai: SCNet: Subdivision Coding Network for Object Detection Based on 3D Point Cloud. IEEE Access 2019.
43 AVOD-FPN
This method makes use of Velodyne laser scans.
 code 58.70 % 69.21 % 53.47 % 0.1 s Titan X (Pascal)
J. Ku, M. Mozifian, J. Lee, A. Harakeh and S. Waslander: Joint 3D Proposal Generation and Object Detection from View Aggregation. IROS 2018.
44 Deep3DBox 58.56 % 68.31 % 50.30 % 1.5 s GPU @ 2.5 Ghz (C/C++)
A. Mousavian, D. Anguelov, J. Flynn and J. Kosecka: 3D Bounding Box Estimation Using Deep Learning and Geometry. CVPR 2017.
45 3DOP
This method uses stereo information.
 code 58.45 % 72.24 % 51.91 % 3s GPU @ 2.5 Ghz (Matlab + C/C++)
X. Chen, K. Kundu, Y. Zhu, A. Berneshawi, H. Ma, S. Fidler and R. Urtasun: 3D Object Proposals for Accurate Object Class Detection. NIPS 2015.
46 Complexer-YOLO
This method makes use of Velodyne laser scans.
 58.28 % 65.41 % 54.27 % 0.06 s GPU @ 3.5 Ghz (C/C++)
M. Simon, K. Amende, A. Kraus, J. Honer, T. Samann, H. Kaulbersch, S. Milz and H. Michael Gross: Complexer-YOLO: Real-Time 3D Object Detection and Tracking on Semantic Point Clouds. The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops 2019.
47 DeepStereoOP 56.55 % 69.36 % 49.37 % 3.4 s GPU @ 3.5 Ghz (Matlab + C/C++)
C. Pham and J. Jeon: Robust Object Proposals Re-ranking for Object Detection in Autonomous Driving Using Convolutional Neural Networks. Signal Processing: Image Communiation 2017.
48 Mono3D code 53.96 % 67.33 % 47.91 % 4.2 s GPU @ 2.5 Ghz (Matlab + C/C++)
X. Chen, K. Kundu, Z. Zhang, H. Ma, S. Fidler and R. Urtasun: Monocular 3D Object Detection for Autonomous Driving. CVPR 2016.
49 DAMNET code 51.78 % 69.61 % 47.49 % 1 s 1 core @ 2.5 Ghz (C/C++)
50 AVOD
This method makes use of Velodyne laser scans.
 code 51.05 % 64.81 % 45.12 % 0.08 s Titan X (pascal)
J. Ku, M. Mozifian, J. Lee, A. Harakeh and S. Waslander: Joint 3D Proposal Generation and Object Detection from View Aggregation. IROS 2018.
51 BirdNet+
This method makes use of Velodyne laser scans.
 code 50.94 % 69.92 % 47.01 % 0.1 s Titan Xp (PyTorch)
A. Barrera, C. Guindel, J. Beltrán and F. García: BirdNet+: End-to-End 3D Object Detection in LiDAR Bird's Eye View. arXiv:2003.04188 [cs.CV] 2020.
52 Mono3CN 50.58 % 66.58 % 45.21 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
53 FRCNN+Or code 49.53 % 63.45 % 43.65 % 0.09 s Titan Xp GPU
C. Guindel, D. Martin and J. Armingol: Fast Joint Object Detection and Viewpoint Estimation for Traffic Scene Understanding. IEEE Intelligent Transportation Systems Magazine 2018.
C. Guindel, D. Martin and J. Armingol: Joint Object Detection and Viewpoint Estimation using CNN features. IEEE International Conference on Vehicular Electronics and Safety (ICVES) 2017.
54 MonoPSR code 49.32 % 58.63 % 43.05 % 0.2 s GPU @ 3.5 Ghz (Python)
J. Ku*, A. Pon* and S. Waslander: Monocular 3D Object Detection Leveraging Accurate Proposals and Shape Reconstruction. CVPR 2019.
55 BirdNet
This method makes use of Velodyne laser scans.
 45.03 % 62.69 % 41.88 % 0.11 s Titan Xp (Caffe)
J. Beltrán, C. Guindel, F. Moreno, D. Cruzado, F. García and A. Escalera: BirdNet: A 3D Object Detection Framework from LiDAR Information. 2018 21st International Conference on Intelligent Transportation Systems (ITSC) 2018.
56 UDI-mono3D 44.08 % 57.89 % 37.89 % 0.05 s 1 core @ 2.5 Ghz (Python)
57 SparsePool code 43.50 % 59.77 % 39.36 % 0.13 s 8 cores @ 2.5 Ghz (Python)
Z. Wang, W. Zhan and M. Tomizuka: Fusing bird view lidar point cloud and front view camera image for deep object detection. arXiv preprint arXiv:1711.06703 2017.
58 UDI-mono3D 42.21 % 57.16 % 36.30 % 0.05 s 1 core @ 2.5 Ghz (Python)
59 NL_M3D 41.19 % 57.44 % 36.24 % 0.2 s 1 core @ 2.5 Ghz (Python)
60 sensekitti code 41.14 % 47.48 % 35.07 % 4.5 s GPU @ 2.5 Ghz (Python + C/C++)
B. Yang, J. Yan, Z. Lei and S. Li: Craft Objects from Images. CVPR 2016.
61 CG-Stereo
This method uses stereo information.
 40.64 % 60.24 % 35.55 % 0.57 s GeForce RTX 2080 Ti
C. Li, J. Ku and S. Waslander: Confidence Guided Stereo 3D Object Detection with Split Depth Estimation. IROS 2020.
62 MonoPair 39.47 % 53.36 % 33.95 % 0.06 s GPU @ 2.5 Ghz (Python + C/C++)
Y. Chen, L. Tai, K. Sun and M. Li: MonoPair: Monocular 3D Object Detection Using Pairwise Spatial Relationships. The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2020.
63 MTMono3d 39.06 % 55.32 % 31.70 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
64 MP-Mono 38.27 % 53.06 % 31.65 % 0.16 s GPU @ 2.5 Ghz (Python)
65 SS3D_HW 37.68 % 52.40 % 32.33 % 0.4 s GPU @ 2.5 Ghz (Python)
66 Disp R-CNN (velo)
This method uses stereo information.
 code 35.77 % 50.66 % 30.96 % 0.42 s GPU @ 2.5 Ghz (Python + C/C++)
J. Sun, L. Chen, Y. Xie, S. Zhang, Q. Jiang, X. Zhou and H. Bao: Disp R-CNN: Stereo 3D Object Detection via Shape Prior Guided Instance Disparity Estimation. CVPR 2020.
67 Disp R-CNN
This method uses stereo information.
 code 35.76 % 50.64 % 30.95 % 0.42 s GPU @ 2.5 Ghz (Python + C/C++)
J. Sun, L. Chen, Y. Xie, S. Zhang, Q. Jiang, X. Zhou and H. Bao: Disp R-CNN: Stereo 3D Object Detection via Shape Prior Guided Instance Disparity Estimation. CVPR 2020.
68 Shift R-CNN (mono) code 34.77 % 51.95 % 31.10 % 0.25 s GPU @ 1.5 Ghz (Python)
A. Naiden, V. Paunescu, G. Kim, B. Jeon and M. Leordeanu: Shift R-CNN: Deep Monocular 3D Object Detection With Closed-form Geometric Constraints. ICIP 2019.
69 SparsePool code 34.56 % 43.33 % 31.09 % 0.13 s 8 cores @ 2.5 Ghz (Python)
Z. Wang, W. Zhan and M. Tomizuka: Fusing bird view lidar point cloud and front view camera image for deep object detection. arXiv preprint arXiv:1711.06703 2017.
70 KNN-GCNN 34.03 % 39.32 % 31.17 % 0.4 s 1 core @ 2.5 Ghz (Python + C/C++)
71 MMCOM 32.52 % 35.29 % 28.89 % 0.04 s 1 core @ 2.5 Ghz (Python)
72 HWFD 32.51 % 35.23 % 28.94 % 0.21 s one 1080Ti
73 vxnet 32.43 % 47.41 % 28.74 % 1 s 1 core @ 2.5 Ghz (C/C++)
74 Point-GNN
This method makes use of Velodyne laser scans.
 code 32.37 % 36.29 % 29.81 % 0.6 s GPU @ 2.5 Ghz (Python)
W. Shi and R. Rajkumar: Point-GNN: Graph Neural Network for 3D Object Detection in a Point Cloud. CVPR 2020.
75 PP-3D 32.37 % 36.29 % 29.81 % 0.1 s 1 core @ 2.5 Ghz (Python)
76 dgist_multiDetNet 31.84 % 36.92 % 28.02 % 0.08 s GPU Titanx Pascal (Python)
77 D4LCN code 31.70 % 48.03 % 26.99 % 0.2 s GPU @ 2.5 Ghz (Python + C/C++)
M. Ding, Y. Huo, H. Yi, Z. Wang, J. Shi, Z. Lu and P. Luo: Learning Depth-Guided Convolutions for Monocular 3D Object Detection. CVPR 2020.
78 Center3D 31.68 % 51.51 % 28.45 % 0.05 s GPU @ 3.5 Ghz (Python)
79 Faster RCNN + Gr + A 31.55 % 36.35 % 28.43 % 1.29 s GPU @ 2.5 Ghz (Python)
80 M3D-RPN code 31.09 % 48.11 % 26.10 % 0.16 s GPU @ 1.5 Ghz (Python)
G. Brazil and X. Liu: M3D-RPN: Monocular 3D Region Proposal Network for Object Detection . ICCV 2019 .
81 Faster RCNN + A 30.81 % 36.25 % 27.51 % 0.19 s GPU @ 2.5 Ghz (Python)
82 Faster RCNN + G 30.61 % 36.19 % 27.22 % 1.1 s GPU @ 2.5 Ghz (Python)
83 LZnet 30.42 % 34.59 % 27.83 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
84 Multi-task DG 30.31 % 35.07 % 26.90 % 0.06 s GPU @ 2.5 Ghz (Python)
85 Faster RCNN + A 30.12 % 36.03 % 26.98 % 0.19 s GPU @ 2.5 Ghz (Python + C/C++)
86 HR-faster-rcnn 29.82 % 36.89 % 26.45 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
87 DP3D 28.41 % 42.17 % 24.02 % 0.07 s GPU @ 1.5 Ghz (Python + C/C++)
88 SS3D 27.79 % 42.95 % 24.26 % 48 ms Tesla V100 (Python)
E. Jörgensen, C. Zach and F. Kahl: Monocular 3D Object Detection and Box Fitting Trained End-to-End Using Intersection-over-Union Loss. CoRR 2019.
89 DP3D 27.47 % 40.80 % 24.16 % 0.05 s GPU @ 1.5 Ghz (Python + C/C++)
90 PG-MonoNet 26.37 % 35.44 % 23.38 % 0.19 s GPU @ 2.5 Ghz (Python)
91 GA2500 26.08 % 32.91 % 22.06 % 0.2 s 1 core @ 2.5 Ghz (Python)
92 GA_rpn500 26.08 % 32.91 % 22.06 % 1 s 1 core @ 2.5 Ghz (Python)
93 DAM 26.05 % 34.25 % 22.30 % 1 s GPU @ 2.5 Ghz (Python)
94 GA_FULLDATA 25.80 % 33.35 % 22.70 % 1 s 4 cores @ 2.5 Ghz (Python)
95 GA_BALANCE 25.27 % 33.79 % 22.03 % 1 s 1 core @ 2.5 Ghz (Python)
96 PMN 25.14 % 31.98 % 21.92 % 0.2 s 1 core @ 2.5 Ghz (Python)
97 bigger_ga 24.64 % 31.31 % 21.06 % 1 s 1 core @ 2.5 Ghz (Python)
98 LAPNet 24.40 % 38.54 % 20.17 % 0.03 s 1 core @ 2.5 Ghz (Python)
99 yolo4_5l 23.96 % 31.36 % 21.02 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
100 PB3D
This method uses stereo information.
 23.93 % 38.00 % 21.58 % 0.42 s 1 core @ 2.5 Ghz (C/C++)
101 CRCNNA 23.88 % 29.91 % 20.70 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
102 GAFM 22.84 % 31.62 % 19.88 % 0.5 s 1 core @ 2.5 Ghz (Python)
103 JSU-NET 22.83 % 31.58 % 19.81 % 0.1 s 1 core @ 2.5 Ghz (Python)
104 yolo4 22.04 % 30.58 % 19.33 % 0.02 s 1 core @ 2.5 Ghz (Python)
105 ga50 21.59 % 29.77 % 18.77 % 1 s 1 core @ 2.5 Ghz (Python)
106 Mag 21.08 % 28.60 % 18.58 % 0.07 s 1 core @ 2.5 Ghz (C/C++)
107 yolo4_5l code 20.79 % 28.67 % 18.35 % 0.02 s 1 core @ 2.5 Ghz (Python + C/C++)
108 DSGN
This method uses stereo information.
 code 20.28 % 29.76 % 19.13 % 0.67 s NVIDIA Tesla V100
Y. Chen, S. Liu, X. Shen and J. Jia: DSGN: Deep Stereo Geometry Network for 3D Object Detection. CVPR 2020.
109 CDI3D 20.12 % 24.76 % 18.37 % 0.03 s GPU @ 2.5 Ghz (Python)
110 I2BEV 19.96 % 30.35 % 17.38 % 0.63 s GPU @ 2.5 Ghz (Python)
111 LSVM-MDPM-sv 19.15 % 26.05 % 18.02 % 10 s 4 cores @ 3.0 Ghz (C/C++)
P. Felzenszwalb, R. Girshick, D. McAllester and D. Ramanan: Object Detection with Discriminatively Trained Part-Based Models. PAMI 2010.
A. Geiger, C. Wojek and R. Urtasun: Joint 3D Estimation of Objects and Scene Layout. NIPS 2011.
112 OC Stereo
This method uses stereo information.
 code 18.99 % 29.07 % 16.40 % 0.35 s 1 core @ 2.5 Ghz (Python + C/C++)
A. Pon, J. Ku, C. Li and S. Waslander: Object-Centric Stereo Matching for 3D Object Detection. ICRA 2020.
113 DPM-VOC+VP 18.92 % 27.97 % 17.43 % 8 s 1 core @ 2.5 Ghz (C/C++)
B. Pepik, M. Stark, P. Gehler and B. Schiele: Multi-view and 3D Deformable Part Models. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) 2015.
114 Scan_YOLO 18.63 % 27.15 % 16.42 % 0.1 s 4 cores @ 3.0 Ghz (Python)
115 yolo_rgb 17.93 % 26.18 % 16.30 % 0.07 s GPU @ 2.5 Ghz (Python)
116 BdCost+DA+BB+MS 17.73 % 23.48 % 14.67 % TBD s 4 cores @ 2.5 Ghz (C/C++)
117 RefinedMPL 16.02 % 26.54 % 13.20 % 0.15 s GPU @ 2.5 Ghz (Python + C/C++)
J. Vianney, S. Aich and B. Liu: RefinedMPL: Refined Monocular PseudoLiDAR for 3D Object Detection in Autonomous Driving. arXiv preprint arXiv:1911.09712 2019.
118 yolo_depth 15.96 % 21.45 % 14.21 % 0.07 s GPU @ 2.5 Ghz (Python)
119 DPM-C8B1
This method uses stereo information.
 14.64 % 23.93 % 13.09 % 15 s 4 cores @ 2.5 Ghz (Matlab + C/C++)
J. Yebes, L. Bergasa and M. García-Garrido: Visual Object Recognition with 3D-Aware Features in KITTI Urban Scenes. Sensors 2015.
J. Yebes, L. Bergasa, R. Arroyo and A. Lázaro: Supervised learning and evaluation of KITTI's cars detector with DPM. IV 2014.
120 EACV 14.32 % 20.90 % 12.56 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
121 BdCost+DA+BB 13.30 % 17.22 % 11.04 % TBD s 4 cores @ 2.5 Ghz (C/C++)
122 RT3D-GMP
This method uses stereo information.
 8.32 % 11.73 % 7.24 % 0.06 s GPU @ 2.5 Ghz (Python + C/C++)
123 RT3DStereo
This method uses stereo information.
 3.88 % 5.46 % 3.54 % 0.08 s GPU @ 2.5 Ghz (C/C++)
H. Königshof, N. Salscheider and C. Stiller: Realtime 3D Object Detection for Automated Driving Using Stereo Vision and Semantic Information. Proc. IEEE Intl. Conf. Intelligent Transportation Systems 2019.
124 Simple3D Net 0.71 % 0.77 % 0.69 % 0.02 s GPU @ 2.5 Ghz (Python)
125 CBNet 0.18 % 0.11 % 0.21 % 1 s 4 cores @ 2.5 Ghz (Python)
126 UM3D_TUM 0.00 % 0.00 % 0.00 % 0.05 s 1 core @ 2.5 Ghz (Python)
Table as LaTeX | Only published Methods


Related Datasets

Citation

When using this dataset in your research, we will be happy if you cite us:
@INPROCEEDINGS{Geiger2012CVPR,
  author = {Andreas Geiger and Philip Lenz and Raquel Urtasun},
  title = {Are we ready for Autonomous Driving? The KITTI Vision Benchmark Suite},
  booktitle = {Conference on Computer Vision and Pattern Recognition (CVPR)},
  year = {2012}
}



eXTReMe Tracker