DeID-VC: Speaker De-identification via Zero-shot Pseudo Voice Conversion.
Ruibin YuanYuxuan WuJacob LiJaxter KimPublished in: INTERSPEECH (2022)
Keyphrases
- synthesized speech
- prosodic features
- text to speech
- speaker verification
- speech recognition
- emotion recognition
- mel frequency cepstral coefficients
- speech synthesis
- speech sounds
- speaker recognition
- audio visual
- automatic speech recognition
- vc dimension
- speaker identification
- speaker diarization
- neural network
- voice recognition
- speaker dependent
- voice and data services
- gaussian mixture model
- feature set
- principal component analysis
- upper bound
- probabilistic model
- feature extraction
- case study
- computer vision
- learning algorithm