Deep-blur: Blind identification and deblurring with convolutional neural networks
{{output}}
We propose a neural network architecture and a training procedure to estimate blurring operators and deblur images from a single degraded image. Our key assumption is that the forward operators can be parameterized by a low-dimensional vector. The models we co... ...
