Self-Attention Channel Combinator Frontend for End-to-End Multichannel Far-Field Speech Recognition.
Rong GongCarl QuillenDushyant SharmaAndrew GoderreJosé LaínezLjubomir MilanovicPublished in: Interspeech (2021)
Keyphrases
- end to end
- speech recognition
- multi channel
- rate allocation
- hidden markov models
- single channel
- automatic speech recognition
- packet loss rate
- language model
- speech signal
- speech processing
- speech recognition systems
- speech synthesis
- congestion control
- speaker identification
- pattern recognition
- speech recognizer
- multipath
- noisy environments
- speech recognition technology
- scalable video
- speaker independent
- linear prediction
- computer vision
- wavelet coefficients
- end to end distortion
- video sequences