PROJECT TITLE :

Beyond A Gaussian Denoiser: Residual Learning Of Deep Cnn For Image Denoising - 2017

ABSTRACT:

The discriminative model learning for image denoising has been recently attracting considerable attentions due to its favorable denoising performance. In this paper, we tend to take one step forward by investigating the construction of feed-forward denoising convolutional neural networks (DnCNNs) to embrace the progress in terribly deep architecture, learning algorithm, and regularization technique into image denoising. Specifically, residual learning and batch normalization are utilised to hurry up the coaching method and boost the denoising performance. Totally different from the present discriminative denoising models that typically train a selected model for additive white Gaussian noise at a sure noise level, our DnCNN model is ready to handle Gaussian denoising with unknown noise level (i.e., blind Gaussian denoising). With the residual learning strategy, DnCNN implicitly removes the latent clean image within the hidden layers. This property motivates us to coach a single DnCNN model to tackle with several general image denoising tasks, like Gaussian denoising, single image super-resolution, and JPEG image deblocking. Our extensive experiments demonstrate that our DnCNN model will not only exhibit high effectiveness in several general image denoising tasks, however additionally be efficiently implemented by making the most of GPU computing.