Joint RGB-Pose Based Human Action Recognition for Anomaly Detection Applications.
Federico AngeliniSyed Mohsen NaqviPublished in: FUSION (2019)
Keyphrases
- anomaly detection
- action recognition
- human pose
- recognizing actions
- static images
- human actions
- intrusion detection
- human activities
- activity recognition
- action detection
- anomalous behavior
- detecting anomalies
- human detection
- network intrusion detection
- computer vision
- human object interactions
- motion history images
- network traffic
- pose estimation
- behavior analysis
- action classification
- body parts
- intrusion detection system
- recognition of human actions
- one class support vector machines
- unsupervised learning
- detect anomalies
- recognizing human actions
- data mining
- action recognition in videos
- human subjects
- co occurrence
- color information
- human body
- visual features
- semi supervised
- d objects
- feature extraction
- feature selection