Multi-Channel Deep Clustering: Discriminative Spectral and Spatial Embeddings for Speaker-Independent Speech Separation.
Zhong-Qiu WangJonathan Le RouxJohn R. HersheyPublished in: ICASSP (2018)
Keyphrases
- multi channel
- speaker independent
- speech recognition
- speaker dependent
- speaker adaptation
- speaker identification
- hidden markov models
- digit recognition
- unsupervised learning
- acoustic models
- speech recognizer
- automatic speech recognition
- k means
- speech signal
- clustering algorithm
- pattern recognition
- single channel
- speech recognition systems
- spectral features
- discriminative training
- maximum likelihood
- semi supervised
- high dimensional data
- feature extraction
- noisy environments
- spectral analysis
- dimensionality reduction
- audio visual
- hyperspectral
- gaussian mixture model
- image segmentation
- image processing
- machine learning