Full paper: active learning cuts surgical video annotation effort by ~50%
santiviquez · x · 2026-11-11
The full paper from the University of Pennsylvania team is now live on Springer (IJCARS, published Nov 2, 2026). The framework combines active learning with dual-loss optimization: a foundation model generates temporally consistent CAMs, trained with a weak-supervision loss on video-level tool-presence labels and a mask loss on human-corrected annotations, refining iteratively in a human-in-the-loop fashion. Annotation effort dropped 50% versus fully manual labeling, with no need for large fully annotated datasets upfront—supporting scalable surgical tool segmentation.
Related event: UPenn Team Speeds Up Surgical Video Annotation by 50% with Active Learning(3 posts)→
More from Research
- CoRL researcher unpacks "agents are too slow for robots" into three distinct critiques — tomssilver · 2026-11-11
- CoRL researcher unpacks "agents are too slow for robots" into three distinct critiques — tomssilver · 2026-11-11
- CoRL researcher unpacks "agents are too slow for robots" into three distinct critiques — tomssilver · 2026-11-11
- CoRL researcher unpacks "agents are too slow for robots" into three distinct critiques — tomssilver · 2026-11-11
- CoRL researcher unpacks "agents are too slow for robots" into three distinct critiques — tomssilver · 2026-11-11
- Formal verification could make AI coding faster: Lean proofs as agent guardrails — bigdata · 2026-11-11