0
A team of researchers has developed novel approaches using computer vision and deep learning to resolve the problem of low-level vision in videos caused by rain and night-time conditions, as well as improve the accuracy of 3D human pose estimation in videos.