Unsupervised Deep Learning for Optical Flow Estimation

Recent work has shown that optical flow estimation can be formulated as a supervised learning problem. Moreover, convolutional networks have been successfully applied to this task. However, supervised flow learning is obfuscated by the shortage of labeled training data. As a consequence, existing methods have to turn to large synthetic datasets for easily computer generated ground truth. In this work, we explore if a deep network for flow estimation can be trained without supervision. Using image warping by the estimated flow, we devise a simple yet effective unsupervised method for learning optical flow, by directly minimizing photometric consistency. We demonstrate that a flow network can be trained from endto-end using our unsupervised scheme. In some cases, our results come tantalizingly close to the performance of methods trained with full supervision.

Authors

• Zhe Ren
• Junchi Yan School of Artificial Intelligence & Department of CSE & MoE Lab of AI, Shanghai Jiao Tong University
• Bingbing Ni Shanghai Jiao Tong University
• Bin Liu
• Xiaokang Yang Shanghai Jiao Tong University, China
• Hongyuan Zha Shenzhen Institute of Artificial Intelligence and Robotics for Society The Chinese University of Hong Kong, Shenzhen

Keywords

No keywords are indexed for this paper.