Unified Architecture for Multichannel End-to-End Speech Recognition With Neural Beamforming.
Tsubasa OchiaiShinji WatanabeTakaaki HoriJohn R. HersheyXiong XiaoPublished in: IEEE J. Sel. Top. Signal Process. (2017)
Keyphrases
- end to end
- speech recognition
- network architecture
- hidden markov models
- speech signal
- language model
- speech synthesis
- speech processing
- speech recognition technology
- speech recognizer
- congestion control
- noisy environments
- automatic speech recognition
- speaker identification
- pattern recognition
- neural network
- speaker independent
- speech recognition systems
- speaker dependent
- speech enhancement
- differentiated services
- isolated word
- linear prediction
- multipath
- frequency domain