A DINOv2-based multi-task framework with task-specific low-rank adapters and a spatial attention module reports state-of-the-art joint activity recognition and semantic segmentation on three endoscopic surgery datasets.
Rethinking exemplars for continual semantic segmentation in endoscopy scenes: Entropy-based mini-batch pseudo-replay,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
EndoARSS: Adapting Spatially-Aware Foundation Model for Efficient Activity Recognition and Semantic Segmentation in Endoscopic Surgery
A DINOv2-based multi-task framework with task-specific low-rank adapters and a spatial attention module reports state-of-the-art joint activity recognition and semantic segmentation on three endoscopic surgery datasets.