Narrow Adaptive Regularization of weights for grapheme-to-phoneme conversion.
Keigo KuboSakriani SaktiGraham NeubigTomoki TodaSatoshi NakamuraPublished in: ICASSP (2014)
Keyphrases
- adaptive regularization
- grapheme to phoneme conversion
- parameter selection
- super resolution
- total variation
- image deblurring
- blind image deconvolution
- regularization parameter
- natural language processing
- linear combination
- image restoration
- image denoising
- loss function
- machine learning
- denoising
- high resolution
- feature selection
- least squares
- motion estimation
- regularization term
- video sequences
- language resources
- image processing
- artificial intelligence