Multi-Task Autoencoder for Noise-Robust Speech Recognition.
Haoyi ZhangConggui LiuNakamasa InoueKoichi ShinodaPublished in: ICASSP (2018)
Keyphrases
- speech recognition
- noisy environments
- multi task
- hidden markov models
- multi task learning
- language model
- speech signal
- speaker identification
- speech synthesis
- automatic speech recognition
- learning tasks
- transfer learning
- background noise
- multi class
- speech recognizer
- speech enhancement
- feature selection
- sparse learning
- pattern recognition
- gaussian processes
- speaker independent
- noise reduction
- noise level
- learning problems
- neural network
- spoken language
- signal to noise ratio
- active learning
- high dimensional
- image processing
- machine learning