DAST-Net: Dense visual attention augmented spatio-temporal network for unsupervised video anomaly detection.
Rangachary KommanduriMrinmoy GhoraiPublished in: Neurocomputing (2024)
Keyphrases
- anomaly detection
- visual attention
- network traffic
- spatio temporal
- network anomaly detection
- unsupervised anomaly detection
- detect anomalies
- eye tracking
- visual saliency
- intrusion detection
- network intrusion
- spatial and temporal
- eye movements
- saliency map
- unsupervised learning
- eye tracking data
- detecting anomalies
- space time
- network intrusion detection
- vision system
- intrusion detection system
- intrusion prevention
- salient regions
- one class support vector machines
- visual attention model
- anomalous behavior
- network security
- normal behavior
- image sequences
- video frames
- video sequences
- image quality
- real time
- multimedia
- moving objects
- supervised learning
- negative selection algorithm
- human actions
- video content
- video data
- key frames
- object based visual attention
- abnormal events
- visual input
- video surveillance
- visual information
- self organizing maps
- higher level
- input image
- object recognition
- machine learning