Parts-Based Feature Extraction of Spectrum of Speech Signal Using Non-Negative Matrix Factorization.
Jeong-Won ParkChang-Keun KimKwang-Seok LeeSi-Young KohKang-In HurPublished in: J. Inform. and Commun. Convergence Engineering (2003)
Keyphrases
- speech signal
- feature extraction
- speaker identification
- spectral subtraction
- speech recognition
- cepstral coefficients
- background noise
- automatic speech recognition
- mel frequency cepstral coefficients
- linear prediction
- hidden markov models
- non stationary
- speaker recognition
- vocal tract
- feature vectors
- noisy environments
- blind source separation
- spectral analysis
- hearing aids
- speech quality
- automatic speech recognition systems
- feature set
- principal component analysis
- feature space
- pattern classification
- pattern recognition
- image processing
- extracted features
- power spectrum
- noisy images
- acoustic features
- wavelet transform
- speech enhancement
- feature selection
- computer vision
- neural network