Multi-channel speech enhancement using sparse coding on local time-frequency structures.
Zhiyuan ZhouZhaogui DingWeifeng LiZhiyong WuLongbiao WangQingmin LiaoPublished in: INTERSPEECH (2014)
Keyphrases
- multi channel
- sparse coding
- single channel
- speech enhancement
- natural images
- sparse representation
- image classification
- linear combination
- generative model
- unsupervised learning
- image representation
- frequency domain
- signal processing
- feature space
- denoising
- high dimensional
- visual words
- image patches
- small number
- image features
- independent component analysis
- wiener filter
- noisy environments
- noise reduction
- signal to noise ratio
- input image
- computer vision
- edge detection