3 papers
cs.CV2026
Redefining Instance Matching: A Unified Framework for Part-Aware Matching in Panoptic Segmentation Evaluation
Erik GroÃkopf, Soumya Snigdha Kundu, Hendrik Möller +9
The Panoptic Quality (PQ) metric is the standard for jointly evaluating instance and semantic segmentation. However, its original definition relies on a One-to-One matching between…
cs.CV2026
Instance Awareness of Multi-class Semantic Segmentation Loss Functions
Soumya Snigdha Kundu, Florian Kofler, Marina Ivory +3
Instance-sensitive losses for semantic segmentation such as blob loss and CC loss were designed to address instance imbalance, ensuring small lesions generate the same gradient as…
cs.CV2026
UltraFlwr -- An Efficient Federated Surgical Object Detection Framework
Yang Li, Soumya Snigdha Kundu, Maxence Boels +6
Surgical object detection in laparoscopic videos enables real-time instrument identification for workflow analysis and skills assessment, but training robust models such as You Onl…