Magnitude replacement of real and imaginary modulation spectrum of acoustic spectrograms for noise-robust speech recognition.
Hsin-Ju HsiehJeih-weih HungPublished in: ICCE-TW (2015)
Keyphrases
- speech recognition
- spectral subtraction
- noisy environments
- speech signal
- noisy speech
- speech recognition systems
- speaker independent
- speech recognizers
- automatic speech recognition
- background noise
- acoustic features
- pattern recognition
- speech synthesis
- language model
- hidden markov models
- speech recognizer
- acoustic models
- speech processing
- speaker verification
- speaker identification
- additive noise
- speaker dependent
- sound source
- speech enhancement
- speech understanding
- digit recognition
- speech recognition technology
- cepstral coefficients
- information retrieval
- vocal tract
- power spectrum
- speaker diarization
- linear prediction
- isolated word
- speaker recognition
- speech recognition errors
- signal to noise ratio
- neural network
- audio visual speech recognition