HCiT: Deepfake Video Detection Using a Hybrid Model of CNN features and Vision Transformer.
Bachir KaddarSid Ahmed FezzaWassim HamidoucheZahid AkhtarAbdenour HadidPublished in: VCIP (2021)
Keyphrases
- hybrid model
- false positives
- artificial neural networks
- feature vectors
- feature extraction
- video clips
- feature space
- pattern recognition
- computer vision
- hybrid models
- video sequences
- real time
- feature set
- detection algorithm
- image features
- video content
- video frames
- video data
- classification accuracy
- low level
- long term
- video retrieval
- support vector regression
- back propagation neural network
- data mining