Learned Collaborative Stereo Refinement
This work addresses noise and inaccuracies in disparity maps for stereo vision applications, representing an incremental improvement with a novel hybrid method.
The paper tackles the problem of denoising and refining disparity maps in stereo vision by proposing a learning-based variational network derived from unrolling a proximal gradient method, achieving efficiency demonstrated on Middlebury 2014 and Kitti 2015 benchmarks.
In this work, we propose a learning-based method to denoise and refine disparity maps of a given stereo method. The proposed variational network arises naturally from unrolling the iterates of a proximal gradient method applied to a variational energy defined in a joint disparity, color, and confidence image space. Our method allows to learn a robust collaborative regularizer leveraging the joint statistics of the color image, the confidence map and the disparity map. Due to the variational structure of our method, the individual steps can be easily visualized, thus enabling interpretability of the method. We can therefore provide interesting insights into how our method refines and denoises disparity maps. The efficiency of our method is demonstrated by the publicly available stereo benchmarks Middlebury 2014 and Kitti 2015.