Autoencoder Bottleneck Features with Multi-Task Optimisation for Improved Continuous Dysarthric Speech Recognition.
Zhengjun YueHeidi ChristensenJon BarkerPublished in: INTERSPEECH (2020)
Keyphrases
- speech recognition
- multi task
- speech recognition systems
- hidden markov models
- language model
- multi task learning
- cepstral coefficients
- automatic speech recognition
- speech synthesis
- speech signal
- speech recognizer
- feature space
- speaker identification
- image features
- feature vectors
- reinforcement learning
- feature extraction
- image processing
- feature selection
- multi class
- information extraction