Learning Independent Object Motion From Unlabelled Stereoscopic Videos
Authors:
- 1 June 2019
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 5587-5596
Abstract
We present a system for learning motion maps of independently moving objects from stereo videos. The only annotations used in our system are 2D object bounding boxes which introduce the notion of objects in our system. Unlike prior learning based approaches which have focused on predicting dense optical flow fields and/or depth maps for images, we propose to predict instance specific 3D scene flow maps and instance masks from which we derive a factored 3D motion map for each object instance. Our network takes the 3D geometry of the problem into account which allows it to correlate the input images and distinguish moving objects from static ones. We present experiments evaluating the accuracy of our 3D flow vectors, as well as depth maps and projected 2D optical flow where our jointly learned system outperforms earlier approaches trained for each task independently.Keywords
All Related Versions
This publication has 29 references indexed in Scilit:
- Object scene flow for autonomous vehicles2015
- EpicFlow: Edge-preserving interpolation of correspondences for optical flow2015
- A primal-dual framework for real-time dense RGB-D scene flow2015
- Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation2014
- Piecewise Rigid Scene Flow2013
- Efficient Dense Scene Flow from Sparse or Dense Stereo Data2008
- A Variational Method for Scene Flow Estimation from Stereo Sequences2007
- Visual Modeling with a Hand-Held CameraInternational Journal of Computer Vision, 2004
- Image quality assessment: from error visibility to structural similarityIEEE Transactions on Image Processing, 2004
- Maximizing Rigidity: The Incremental Recovery of 3-D Structure from Rigid and Nonrigid MotionPerception, 1984