Phase-sensitive and recognition-boosted speech separation using deep recurrent neural networks.
Hakan ErdoganJohn R. HersheyShinji WatanabeJonathan Le RouxPublished in: ICASSP (2015)
Keyphrases
- recurrent neural networks
- recognition engine
- neural network
- recognition rate
- echo state networks
- recurrent networks
- feed forward
- complex valued
- recognition accuracy
- speech recognition
- object recognition
- artificial neural networks
- reservoir computing
- automatic speech recognition systems
- pattern recognition
- speech recognition systems
- neural model
- continuous speech recognition
- speech corpus
- spoken words
- long short term memory
- speaker independent
- real time
- artificial intelligence
- chaotic time series
- cascade correlation
- speech signal
- long term