Target speech GMM-based spectral compensation for noise robust speech recognition.
Takahiro ShinozakiSadaoki FuruiPublished in: INTERSPEECH (2009)
Keyphrases
- speech recognition
- noisy environments
- spectral subtraction
- speaker identification
- speech signal
- voice activity detection
- noisy speech
- automatic speech recognition
- speaker recognition
- speech processing
- speech synthesis
- language model
- hidden markov models
- speech enhancement
- background noise
- speaker verification
- pattern recognition
- speech recognizer
- speech recognition systems
- spectral analysis
- speech recognition technology
- cepstral coefficients
- speaker adaptation
- probabilistic neural network
- speaker dependent
- handwriting recognition
- digit recognition
- speaker independent
- linear prediction
- word error rate
- mel frequency cepstral coefficients
- gaussian mixture model
- recognition engine
- keyword spotting
- speech recognition errors
- signal to noise ratio
- computer vision
- noise reduction