Speech enhancement by sparse, low-rank, and dictionary spectrogram decomposition.
Zhuo ChenDaniel P. W. EllisPublished in: WASPAA (2013)
Keyphrases
- low rank
- speech enhancement
- tensor decomposition
- speech signal
- single channel
- low rank matrix
- rank minimization
- low rank matrices
- sparse representation
- noisy environments
- multi channel
- robust principal component analysis
- missing data
- linear combination
- matrix completion
- matrix factorization
- high dimensional data
- convex optimization
- speech recognition
- regularized regression
- noise reduction
- singular value decomposition
- high order
- signal to noise ratio
- non stationary
- automatic speech recognition
- semi supervised
- spectral analysis
- frequency domain
- wiener filter
- high dimensional
- sparse matrix
- hidden markov models
- small number
- nearest neighbor
- markov random field
- signal processing
- low dimensional
- recommender systems