Speech Separation based on signal-noise-dependent deep neural networks for robust speech recognition.
Yanhui TuJun DuLi-Rong DaiChin-Hui LeePublished in: ICASSP (2015)
Keyphrases
- speech recognition
- noisy environments
- noisy speech
- spectral subtraction
- speech signal
- neural network
- pattern recognition
- speech enhancement
- automatic speech recognition
- background noise
- hidden markov models
- speech synthesis
- speaker verification
- speaker identification
- speech recognizer
- language model
- digit recognition
- speech recognition systems
- speech processing
- speaker independent
- speech recognition technology
- keyword spotting
- additive noise
- signal processing
- sound source
- speech recognizers
- noise reduction
- speech retrieval
- voice activity detection
- speech recognition errors
- recognition engine
- vocal tract
- handwriting recognition
- speaker dependent
- speaker diarization
- word recognition
- audio visual speech recognition