Content-Insensitive Dynamic Lip Feature Extraction for Visual Speaker Authentication Against Deepfake Attacks.
Zihao GuoShilin WangPublished in: ICASSP (2023)
Keyphrases
- feature extraction
- dynamic content
- security mechanisms
- visual speech
- countermeasures
- semi fragile watermarking
- visual information
- multimedia
- face recognition
- preprocessing
- image classification
- detect malicious
- multimedia content
- watermarking scheme
- feature vectors
- wavelet transform
- semantic content
- watermarking algorithm
- image processing
- image authentication
- speaker recognition
- feature space
- discriminant analysis
- low level
- speech recognition
- principal component analysis