ModuSeg enables training-free weakly supervised semantic segmentation by explicitly separating geometric object discovery from non-parametric semantic feature retrieval using existing mask proposers and foundation model feature banks.
In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
ModuSeg: Decoupling Object Discovery and Semantic Retrieval for Training-Free Weakly Supervised Segmentation
ModuSeg enables training-free weakly supervised semantic segmentation by explicitly separating geometric object discovery from non-parametric semantic feature retrieval using existing mask proposers and foundation model feature banks.