Speaker anonymization by modifying fundamental frequency and x-vector singular value.
Candy Olivia MawalimKasorn GalajitJessada KarnjanaShunsuke KidaniMasashi UnokiPublished in: Comput. Speech Lang. (2022)
Keyphrases
- singular values
- fundamental frequency
- singular value decomposition
- speech signal
- speech recognition
- feature vectors
- information loss
- privacy preserving
- fundamental matrix
- low rank
- automatic speech recognition
- speaker verification
- neural network
- viewpoint
- low dimensional
- audio visual
- pairwise
- data matrix
- acoustic features
- computer vision