Using an autoencoder with deformable templates to discover features for automated speech recognition.
Navdeep JaitlyGeoffrey E. HintonPublished in: INTERSPEECH (2013)
Keyphrases
- speech recognition
- deformable templates
- speech recognition systems
- hidden markov models
- cepstral coefficients
- speech recognizer
- speech signal
- speech synthesis
- pattern recognition
- noisy environments
- automatic speech recognition
- language model
- speech processing
- feature set
- feature extraction
- classification accuracy
- mel frequency cepstral coefficients
- low level
- speech recognition technology
- co occurrence
- speaker identification
- support vector machine
- feature space
- speaker dependent