Compositional embedding models for speaker identification and diarization with simultaneous speech from 2+ speakers.
Zeqian LiJacob WhitehillPublished in: CoRR (2020)
Keyphrases
- speaker identification
- speech recognition
- speaker dependent
- speech signal
- gaussian mixture model
- speaker recognition
- speech processing
- speaker diarization
- noisy environments
- feature extraction
- broadcast news
- automatic speech recognition
- probabilistic model
- hidden markov models
- speaker independent
- model selection
- language model
- edge detection
- speech recognition systems
- speaker adaptation
- machine learning