Speech Enhancement by Spectral Subtraction Based on Subspace Decomposition.
Takahiro MurakamiTetsuya HoyaYoshihisa IshidaPublished in: IEICE Trans. Fundam. Electron. Commun. Comput. Sci. (2005)
Keyphrases
- spectral subtraction
- speech enhancement
- speech signal
- speech recognition
- hybrid method
- noise reduction
- noisy environments
- signal to noise ratio
- noisy speech
- single channel
- linear prediction
- hidden markov models
- vocal tract
- smoothing algorithm
- wiener filter
- background noise
- feature space
- automatic speech recognition
- principal component analysis
- high dimensional
- hearing aids
- multi channel
- hybrid algorithm
- non stationary
- computer vision
- independent component analysis
- markov random field
- denoising
- image processing