Breaking trade-offs in speech separation with sparsely-gated mixture of experts.
Xiaofei WangZhuo ChenYu ShiJian WuNaoyuki KandaTakuya YoshiokaPublished in: CoRR (2022)
Keyphrases
- trade off
- blind separation
- speech recognition
- mixture model
- gaussian mixture model
- speech signal
- speaker recognition
- automatic speech recognition
- audio visual
- endpoint detection
- speech synthesis
- text to speech
- sparse representation
- expectation maximization
- gaussian distribution
- blind source separation
- language acquisition
- expert finding
- speaker identification
- spoken dialogue systems
- feature space
- recognition engine
- neural network
- human experts
- multi modal
- pattern recognition
- fundamental frequency