Noise reduction for noise robust feature extraction for distributed speech recognition.
Bernhard NoéJürgen SienelDenis JouvetLaurent MauuaryJohan de VethLou BovesFebe de WetPublished in: INTERSPEECH (2001)
Keyphrases
- noisy environments
- noise reduction
- speech recognition
- speaker identification
- feature extraction
- signal to noise ratio
- speech enhancement
- edge preserving
- background noise
- noise level
- noise removal
- edge detection
- pattern recognition
- median filter
- noise free
- noise detection
- speech signal
- wiener filter
- automatic speech recognition
- hidden markov models
- word recognition
- language model
- speech recognizer
- speech synthesis
- frequency domain
- principal component analysis
- cepstral coefficients
- spectral subtraction
- speech recognition technology
- noisy speech
- extracting features
- wavelet transform
- feature vectors
- image processing
- speaker recognition
- impulse noise
- speech recognition systems
- information retrieval