研究室发布CVPR 2026视频生成论文NS-Diff源代码
Zijun Deng and Yuxin Peng*, "NS-Diff: Fluid Navier–Stokes Guided Video Diffusion via Reinforcement Learning", The IEEE/CVF Conference on Computer Vision and Pattern Recognition 2026 (CVPR), Denver CO, USA, Jun. 3-7, 2026. 【pdf】【source code】【CVPR海报】【机器之心报道】