AI RESEARCH
Retrieve What's Missing: Coverage-Maximizing Retrieval for Consistent Long Video Generation
arXiv CS.CV
•
ArXi:2606.02479v1 Announce Type: new Maintaining long-term geometric consistency remains challenging for long-horizon autoregressive video generation. Memory-augmented generative models address this by retrieving historical frames, but their effectiveness depends on two key design choices: what 3D-geometric evidence should represent past observations, and how memory frames should be selected from this evidence.