Acoustic feature compensation based on decomposition of speech and noise for ASR in noisy environments.
Hong Kook KimRichard C. RoseHong-Goo KangPublished in: INTERSPEECH (2001)
Keyphrases
- noisy environments
- automatic speech recognition
- noisy speech
- speech recognition
- speech enhancement
- speech signal
- acoustic features
- noise reduction
- speaker verification
- speech sounds
- background noise
- speech recognition systems
- speech recognizers
- speaker identification
- digit recognition
- voice activity detection
- visual speech
- vocal tract
- speech recognizer
- word error rate
- word recognition
- language model
- hidden markov models
- spectral subtraction
- class noise
- feature vectors
- mel frequency cepstral coefficients
- speaker recognition
- sound source
- linear prediction
- formant frequencies
- speech synthesis
- broadcast news
- feature selection