Distinguishable Speaker Anonymization Based on Formant and Fundamental Frequency Scaling.
Jixun YaoQing WangYi LeiPengcheng GuoLei XieNamin WangJie LiuPublished in: ICASSP (2023)
Keyphrases
- fundamental frequency
- speech signal
- speech recognition
- automatic speech recognition
- speaker identification
- vocal tract
- speaker recognition
- acoustic features
- privacy preserving
- noisy environments
- information loss
- hidden markov models
- privacy protection
- speaker diarization
- privacy preservation
- pattern recognition
- speaker adaptation
- speaker verification
- data privacy
- non stationary
- computer vision
- broadcast news
- audio visual
- preserving privacy
- privacy preserving data publishing
- differential privacy
- maximum likelihood
- language model
- speech synthesis
- feature space
- image processing
- neural network