Using Self Attention DNNs to Discover Phonemic Features for Audio Deep Fake Detection.
Hira DhamyalAyesha AliIhsan Ayyub QaziAgha Ali RazaPublished in: ASRU (2021)
Keyphrases
- false positives
- feature extraction
- support vector machine classifier
- feature space
- low level
- multimedia
- detection accuracy
- audio visual
- false alarms
- automatic detection
- anomaly detection
- feature set
- image features
- classification accuracy
- hidden markov models
- feature vectors
- object recognition
- co occurrence
- image classification
- event detection
- feature detection
- multiscale
- neural network
- adaboost classifier
- cepstral features