FedVAR aggregates per-client normality prototypes into one global anchor and re-centers CLIP visual and text features around it, improving federated fine-grained video anomaly recognition.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
FedVAR: Prototype-Aligned Federated Framework for Video Anomaly Recognition
FedVAR aggregates per-client normality prototypes into one global anchor and re-centers CLIP visual and text features around it, improving federated fine-grained video anomaly recognition.