Unsupervised Moving Object Detection via Contextual Information Separation
- 1 June 2019
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 879-888
Abstract
We propose an adversarial contextual model for detecting moving objects in images. A deep neural network is trained to predict the optical flow in a region using information from everywhere else but that region (context), while another network attempts to make such context as uninformative as possible. The result is a model where hypotheses naturally compete with no need for explicit regularization or hyper-parameter tuning. Although our method requires no supervision whatsoever, it outperforms several methods that are pre-trained on large annotated datasets. Our model can be thought of as a generalization of classical variational generative region-based segmentation, but in a way that avoids explicit regularization or solution of partial differential equations at run-time.Keywords
All Related Versions
This publication has 33 references indexed in Scilit:
- Towards the quantitative evaluation of visual attention modelsVision Research, 2015
- Shape Tracking with Occlusions via Coarse-to-Fine Region-Based Sobolev DescentIEEE Transactions on Pattern Analysis and Machine Intelligence, 2014
- Segmentation of Moving Objects by Long Term Video AnalysisIEEE Transactions on Pattern Analysis and Machine Intelligence, 2013
- A Fully-Connected Layered Model of Foreground and Background Flow2013
- Medical Image Segmentation: A Brief Survey2011
- Variational Motion Segmentation with Level Sets2006
- Motion Competition: A Variational Approach to Piecewise Parametric Motion SegmentationInternational Journal of Computer Vision, 2004
- A real-time computer vision system for measuring traffic parameters2002
- Active contours without edgesIEEE Transactions on Image Processing, 2001
- Representing moving images with layersIEEE Transactions on Image Processing, 1994