Speaker normalization with a mixture of recurrent networks.
Edmondo TrentinDiego GiulianiPublished in: ESANN (1997)
Keyphrases
- recurrent networks
- recurrent neural networks
- biologically inspired
- feed forward
- neural network
- mixture model
- speech recognition
- speaker recognition
- gaussian mixture model
- preprocessing
- audio visual
- gaussian distribution
- normalization method
- speaker verification
- speaker identification
- latent variable models
- logit model
- artificial neural networks
- expectation maximization
- high dimensional
- feature extraction
- image segmentation
- case study
- feature selection
- heavy tailed
- search engine
- artificial intelligence
- speaker diarization
- blind separation
- learning algorithm