In this work, we introduce DINOv, a Visual In-Context Prompting framework for referring and generic segmentation tasks. For visualization and demos, we also recommend ...