Reinforcement Learning-based Mixture of Vision Transformers for Video Violence Recognition.
Hamid MohammadiEhsan NazerfardTahereh FirooziPublished in: CoRR (2023)
Keyphrases
- reinforcement learning
- real time
- recognition accuracy
- video data
- object recognition
- human activities
- computer vision
- video sequences
- recognition rate
- video content
- video streams
- pattern recognition
- high level vision
- video clips
- mixture model
- feature extraction
- multimedia
- space time
- automatic recognition
- recognition process
- model free
- video database
- character recognition
- recognition algorithm
- learning algorithm
- video surveillance
- image processing
- event detection
- text detection
- video processing
- static images
- neural network
- action selection
- video retrieval
- face recognition
- action recognition
- multi agent