Noise reduction and evaluation of speech recognition performance by the adaptive beamformer with four microphone for wearable speech translation device.
Kohhei HayashidaTsuyoki NishikawaKazuya NomuraTakeo KanamoriTakanori AoyamaPublished in: GCCE (2017)
Keyphrases
- speech recognition
- noise reduction
- automatic speech recognition
- noisy environments
- signal to noise ratio
- speech signal
- hearing aids
- speech enhancement
- edge preserving
- speech synthesis
- speaker identification
- hidden markov models
- speech processing
- noisy speech
- speech retrieval
- language model
- spectral subtraction
- speaker diarization
- speech recognizer
- speech recognition systems
- noise level
- speech recognition technology
- word error rate
- speaker verification
- wiener filter
- adaptive filtering
- pattern recognition
- edge detection
- background noise
- speech recognizers
- neural network
- sound source
- vocal tract
- digit recognition
- word recognition
- noise removal
- broadcast news
- cepstral coefficients
- machine translation
- information retrieval
- multiscale
- non stationary
- speaker dependent
- speaker adaptation
- speaker recognition