Combining a mobile deep neural network and a recurrent layer for violence detection in videos.
Paolo ContardoSelene TomassiniNicola FalcionelliAldo Franco DragoniPaolo SernaniPublished in: RTA-CSIT (2023)
Keyphrases
- neural network
- recurrent neural networks
- multi layer
- multi modal fusion
- feed forward
- automatic detection
- multiple layers
- object detection
- detection rate
- detection method
- video sequences
- single layer
- endoscopic video
- detection algorithm
- video indexing
- false positives
- mobile devices
- artificial neural networks
- false alarms
- video analysis
- neural nets
- event detection
- video frames
- anomaly detection
- mobile phone
- genetic algorithm
- dynamic scenes
- multilayer perceptron
- self organizing maps
- back propagation
- video data
- context aware
- neural network is trained