Geometry-based Occlusion-Aware Unsupervised Stereo Matching for Autonomous Driving
This work addresses occlusion handling in stereo matching for autonomous driving, offering an incremental improvement with a novel geometry-based approach that can be extended to other methods.
The paper tackles the occlusion problem in unsupervised stereo matching for autonomous driving by introducing a geometry-based occlusion detection method and a novel training strategy that uses predicted disparity maps as pseudo groundtruth, resulting in significant performance improvements over other unsupervised methods.
Recently, there are emerging many stereo matching methods for autonomous driving based on unsupervised learning. Most of them take advantage of reconstruction losses to remove dependency on disparity groundtruth. Occlusion handling is a challenging problem in stereo matching, especially for unsupervised methods. Previous unsupervised methods failed to take full advantage of geometry properties in occlusion handling. In this paper, we introduce an effective way to detect occlusion regions and propose a novel unsupervised training strategy to deal with occlusion that only uses the predicted left disparity map, by making use of its geometry features in an iterative way. In the training process, we regard the predicted left disparity map as pseudo groundtruth and infer occluded regions using geometry features. The resulting occlusion mask is then used in either training, post-processing, or both of them as guidance. Experiments show that our method could deal with the occlusion problem effectively and significantly outperforms the other unsupervised methods for stereo matching. Moreover, our occlusion-aware strategies can be extended to the other stereo methods conveniently and improve their performances.