Speech enhancement and recognition using multi-task learning of long short-term memory recurrent neural networks.
Zhuo ChenShinji WatanabeHakan ErdoganJohn R. HersheyPublished in: INTERSPEECH (2015)
Keyphrases
- recurrent neural networks
- multi task learning
- long short term memory
- speech enhancement
- noisy environments
- noise reduction
- multi task
- signal to noise ratio
- speech signal
- single channel
- neural network
- learning tasks
- linear prediction
- feed forward
- gaussian processes
- learning problems
- echo state networks
- high order
- theoretical analysis
- transfer learning
- learning models
- artificial neural networks
- learning algorithm
- speech recognition
- inductive learning
- multiscale
- multi class
- multi channel
- machine learning algorithms