Synth2Aug: Cross-Domain Speaker Recognition with TTS Synthesized Speech.
Yiling HuangYutian ChenJason PelecanosQuan WangPublished in: SLT (2021)
Keyphrases
- cross domain
- speaker recognition
- synthesized speech
- gaussian mixture model
- vector quantization
- speaker verification
- knowledge transfer
- speaker identification
- probabilistic neural network
- text categorization
- transfer learning
- machine learning
- mixture model
- speech recognition
- image compression
- feature vectors
- feature selection