SENSE-VAD introduces the first synthetic benchmark dataset with per-frame labels for socially complex anomalies in autonomous driving scenes and shows existing video anomaly detectors fail on them.
Jones, Yasin Yilmaz, and Anoop Cherian
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
citation-role summary
baseline 1
citation-polarity summary
fields
cs.CV 2years
2026 2verdicts
UNVERDICTED 2roles
baseline 1polarities
baseline 1representative citing papers
Video anomaly detection is misframed by multi-scene LLM models that reduce the task to semantic action recognition instead of capturing local scene normality, requiring a return to single-scene spatially-aware methods.
citing papers explorer
-
SENSE-VAD: Sentient and Semantic Video Anomaly Detection for Autonomous Driving
SENSE-VAD introduces the first synthetic benchmark dataset with per-frame labels for socially complex anomalies in autonomous driving scenes and shows existing video anomaly detectors fail on them.
-
Is Video Anomaly Detection Misframed? Evidence from LLM-Based and Multi-Scene Models
Video anomaly detection is misframed by multi-scene LLM models that reduce the task to semantic action recognition instead of capturing local scene normality, requiring a return to single-scene spatially-aware methods.