Modelpedia
Work in progress
About
Findings
Models
Concepts
Methods
Datasets
Sources
Light
Dark
The Devil is in the Object Boundary: Towards Annotation-free Instance Segmentation using Foundation Models
2024-01-16
· ICLR 2024 poster ·
anchor
Findings
IC-667
CLIP's intermediate layer features encode object boundaries recoverable by k-means clustering, a property absent in shallow and deep layers
IC-668
SAM's edge-oriented segmentation yields high recall but very low precision because it cannot distinguish object boundaries from interior edges
IC-669
DINOv2's features, when clustered, produce smooth semantic regions but lack instance-level boundary delineation