Just Go With the Flow: Self-Supervised Scene Flow Estimation
Authors:
- 1 June 2020
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 11174-11182
Abstract
When interacting with highly dynamic environments, scene flow allows autonomous systems to reason about the non-rigid motion of multiple independent objects. This is of particular interest in the field of autonomous driving, in which many cars, people, bicycles, and other objects need to be accurately tracked. Current state-of-the-art methods require annotated scene flow data from autonomous driving scenes to train scene flow networks with supervised learning. As an alternative, we present a method of training scene flow that uses two self-supervised losses, based on nearest neighbors and cycle consistency. These self-supervised losses allow us to train our method on large unlabeled autonomous driving datasets; the resulting method matches current state-of-the-art supervised performance using no real world annotations and exceeds state-of-the-art performance when combining our self-supervised approach with supervised learning on a smaller labeled dataset.Keywords
All Related Versions
This publication has 20 references indexed in Scilit:
- A Large Dataset to Train Convolutional Networks for Disparity, Optical Flow, and Scene Flow Estimation2016
- Unsupervised Learning of Visual Representations Using Videos2015
- JOINT 3D ESTIMATION OF VEHICLES AND SCENE FLOWISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 2015
- Object scene flow for autonomous vehicles2015
- 3D Scene Flow Estimation with a Piecewise Rigid Scene ModelInternational Journal of Computer Vision, 2015
- Piecewise Rigid Scene Flow2013
- Multi-view Scene Flow Estimation: A View Centered Variational ApproachInternational Journal of Computer Vision, 2012
- 3D scene flow estimation with a rigid motion prior2011
- Joint Estimation of Motion, Structure and Geometry from Stereo Sequences2010
- Method for registration of 3-D shapes1992