The KITTI Vision Benchmark Suite

We present a deep learning method for end-to-end monocular 3D object detection and metric shape retrieval. We propose a novel loss formulation by lifting 2D detection, orientation, and scale estimation into 3D space. Instead of optimizing these quantities separately, the 3D instantiation allows to properly measure the metric misalignment of boxes.

Lr=0.001
SGD with Momentum

@inproceedings{manhardt2018roi10d,
title={ROI-10D: Monocular Lifting of 2D Detection to 6D Pose and Metric Shape},
author={Manhardt, Fabian and Kehl, Wadim and Gaidon, Adrien},
booktitle={Computer Vision and Pattern Recognition (CVPR)},
year={2019},
organization={IEEE}
}