VESSA: Video-based objEct-centric Self-Supervised Adaptation for Visual Foundation Models

Open in new window