Exploiting Nonnegative Matrix Factorization with Mixed Group Sparsity Constraint to Separate Speech Signal from Single-channel Mixture with Unknown Ambient Noise.
Hien-Thanh DuongCong-Phuong NguyenQuoc-Cuong NguyenPublished in: EAI Endorsed Trans. Context aware Syst. Appl. (2018)
Keyphrases
- speech enhancement
- single channel
- speech signal
- multi channel
- noisy environments
- speech recognition
- noise reduction
- additive noise
- sound source
- wiener filter
- frequency domain
- matrix factorization
- signal to noise ratio
- independent component analysis
- non stationary
- prior information
- hidden markov models
- sparse coding
- machine learning
- higher order
- collaborative filtering
- least squares
- multiresolution
- pattern recognition