Speaker conditioning of acoustic models using affine transformation for multi-speaker speech recognition.
Midia YousefiJohn H. L. HansePublished in: CoRR (2021)
Keyphrases
- speech recognition
- acoustic models
- affine transformation
- speaker independent
- speech recognizer
- hidden markov models
- automatic speech recognition
- pattern recognition
- speaker dependent
- speech signal
- speaker identification
- language model
- broadcast news
- feature points
- speaker diarization
- image matching
- speech synthesis
- b spline
- image registration
- noisy environments
- speaker recognition
- speech recognition systems
- speaker adaptation
- discriminative training
- similarity measure