V-Bridge: Bridging Video Generative Priors to Versatile Few-shot Image Restoration

ArXi:2603.13089v1 Announce Type: new Large-scale video generative models are trained on vast and diverse visual data, enabling them to internalize rich structural, semantic, and dynamic priors of the visual world. While these models have nstrated impressive generative capability, their potential as general-purpose visual learners remains largely untapped. In this work, we