AI RESEARCH
Metric-Guided Feature Fusion of Visual Foundation Models for Segmentation Tasks
arXiv CS.CV
•
ArXi:2605.16864v1 Announce Type: new Although large-scale visual foundation models (VFMs) achieve remarkable performance in semantic understanding, they still underperform in instance-aware dense prediction tasks. They exhibit different biases in representation: for instance, promptable segmentation models (e.g., SAM2) focus on fine-grained region boundaries, while self-supervised models (e.g., DINOv3) emphasize object-level structure. This observation highlights the potential of combining complementary features from different VFMs to enhance downstream dense prediction tasks.