Blind Denoising Autoencoder
This addresses the issue of denoising images when test data differs from training data, which is a common problem in image processing, but it is an incremental improvement over existing blind denoising methods.
The paper tackles the problem of blind denoising by proposing an autoencoder-based method that learns from the noisy sample itself during denoising, unlike prior approaches that require separate training data, and it outperforms existing methods like KSVD, transform learning, sparse stacked denoising autoencoder, and BM3D in experiments.
The term blind denoising refers to the fact that the basis used for denoising is learnt from the noisy sample itself during denoising. Dictionary learning and transform learning based formulations for blind denoising are well known. But there has been no autoencoder based solution for the said blind denoising approach. So far autoencoder based denoising formulations have learnt the model on a separate training data and have used the learnt model to denoise test samples. Such a methodology fails when the test image (to denoise) is not of the same kind as the models learnt with. This will be first work, where we learn the autoencoder from the noisy sample while denoising. Experimental results show that our proposed method performs better than dictionary learning (KSVD), transform learning, sparse stacked denoising autoencoder and the gold standard BM3D algorithm.