Principles and Typical Computational Limitations of Sparse Speaker Separation Based on Deterministic Speech Features.
Albert KernRuedi StoopPublished in: Neural Comput. (2011)
Keyphrases
- speech recognition
- speaker recognition
- audio visual
- automatic speech recognition
- spectral features
- feature vectors
- image features
- neural network
- feature extraction
- high dimensional
- co occurrence
- speaker verification
- canonical correlation analysis
- prosodic features
- text to speech
- audio features
- structural features
- extracted features
- video data
- multi modal
- image classification
- image retrieval
- feature space
- real world