The KITTI Vision Benchmark Suite

Method

Unsupervised Lidar Feature Learning with ESGVI [ULF-ESGVI]

Submitted on 14 Dec. 2020 17:58 by
David Yoon (University of Toronto Institute for Aerospace Studies)

Running time:		0.3 s
Environment:		GPU and CPU @ 2.2 Ghz (Python + C/C++)

Method Description:

Lidar feature learning with a pointcloud convolution network. We learn only from the on-board lidar data, no groundtruth features or trajectory. Inference is sliding-window trajectory optimization with Gauss-Newton. No loop closures.

Parameters:

See paper.

Latex Bibtex:

@article{yoon_ral21,
Author = {D J Yoon and H Zhang and M Gridseth and H Thomas and T D Barfoot},
Doi = {10.1109/LRA.2021.3060407},
Journal = {IEEE Robotics and Automation Letters (RAL)},
Note = { , to be presented at ICRA 2021.},
Title = {Unsupervised Learning of Lidar Features for Use in a Probabilistic Trajectory Estimator},
Year = {2021}}

Detailed Results

From all test sequences (sequences 11-21), our benchmark computes translational and rotational errors for all possible subsequences of length (5,10,50,100,150,...,400) meters. Our evaluation ranks methods according to the average of those values, where errors are measured in percent (for translation) and in degrees per meter (for rotation). Details for different trajectory lengths and driving speeds can be found in the plots underneath. Furthermore, the first 5 test trajectories and error plots are shown below.