AI RESEARCH
INSID3: Training-Free In-Context Segmentation with DINOv3
arXiv CS.CV
•
ArXi:2603.28480v1 Announce Type: new In-context segmentation (ICS) aims to segment arbitrary concepts, e.g., objects, parts, or personalized instances, given one annotated visual examples. Existing work relies on (i) fine-tuning vision foundation models (VFMs), which improves in-domain results but harms generalization, or (ii) combines multiple frozen VFMs, which preserves generalization but yields architectural complexity and fixed segmentation granularities.