A patch-based spatiotemporal localization framework for weakly supervised video anomaly detection that uses grid-level features and a proximity-aware top-k selection strategy to produce fine-grained anomaly maps.
Weakly-supervised spatio-temporal anomaly detection in surveillance video
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
-
Learning Where and When: Patch-Based Spatiotemporal Localization in Weakly Supervised Video Anomaly Detection
A patch-based spatiotemporal localization framework for weakly supervised video anomaly detection that uses grid-level features and a proximity-aware top-k selection strategy to produce fine-grained anomaly maps.