My research focuses on design of efficient algorithm for image/video processing with Transformer-like models. Recently interested in 3D tasks, too.