Blind Denoising Autoencoder

TL;DR

This paper introduces a novel autoencoder-based approach for blind denoising that learns directly from noisy samples during the denoising process, outperforming existing dictionary and transform learning methods.

Contribution

First autoencoder-based blind denoising method that learns from noisy samples during denoising, eliminating the need for separate training data.

Findings

Outperforms dictionary learning (KSVD) and transform learning methods.

Achieves better results than sparse stacked denoising autoencoder.

Surpasses the performance of the BM3D algorithm.

Abstract

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.