Gradient-Based Deep Quantization of Neural Networks through Sinusoidal Adaptive Regularization.
Ahmed T. ElthakebPrannoy PilligundlaFatemehsadat MireshghallahTarek ElgindiCharles-Alban DeledalleHadi EsmaeilzadehPublished in: CoRR (2020)
Keyphrases
- adaptive regularization
- neural network
- parameter selection
- super resolution
- total variation
- image deblurring
- regularization parameter
- blind image deconvolution
- neural network model
- artificial neural networks
- image restoration
- high frequency
- denoising
- computational complexity
- radial basis function
- higher order
- high resolution
- multiscale
- high quality