Noise2Kernel: Adaptive Self-Supervised Blind Denoising using a Dilated Convolutional Kernel Architecture
This work is significant for researchers and practitioners in image processing, offering an incremental improvement in blind denoising, especially for images corrupted by non-zero-mean or hybrid noise.
This paper addresses the problem of blind image denoising, particularly for extreme noise like salt-and-pepper, where existing methods suffer from brightness shifting due to assumptions of zero-mean, signal-independent noise. The authors propose a dilated convolutional network with an adaptive self-supervision loss that circumvents these assumptions, demonstrating efficacy against state-of-the-art methods.
With the advent of recent advances in unsupervised learning, efficient training of a deep network for image denoising without pairs of noisy and clean images has become feasible. However, most current unsupervised denoising methods are built on the assumption of zero-mean noise under the signal-independent condition. This assumption causes blind denoising techniques to suffer brightness shifting problems on images that are greatly corrupted by extreme noise such as salt-and-pepper noise. Moreover, most blind denoising methods require a random masking scheme for training to ensure the invariance of the denoising process. In this paper, we propose a dilated convolutional network that satisfies an invariant property, allowing efficient kernel-based training without random masking. We also propose an adaptive self-supervision loss to circumvent the requirement of zero-mean constraint, which is specifically effective in removing salt-and-pepper or hybrid noise where a prior knowledge of noise statistics is not readily available. We demonstrate the efficacy of the proposed method by comparing it with state-of-the-art denoising methods using various examples.