Taming generative video models for zero-shot optical flow extraction Paper • 2507.09082 • Published Jul 11 • 12
Self-Supervised Learning of Motion Concepts by Optimizing Counterfactuals Paper • 2503.19953 • Published Mar 25 • 3