The KITTI Vision Benchmark Suite

Method

3D-Geometric constrained Deep Visual Odometry [3DG-DVO]
[Anonymous Submission]

Submitted on 5 Jun. 2024 20:30 by
[Anonymous Submission]

Running time:		0.04 s
Environment:		GPU @ 1.5 Ghz (Python)

Method Description:

A simple but highly effective self-supervised learning framework for deep visual odometry on stereo cameras.
Its central component is a deep neural network for optical flow predictions, from which depth and ego-motion can be derived.
The optical flow network is guided by a 3D-geometric constraint, which enforces a realistic structure of the scene over consecutive frames and models static and moving objects.

Parameters:

Parameters will be reported in the ICLR24 paper.

Latex Bibtex:

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.