In the second stage, we use a refinement network learned with adversarial training for further improvement of the image quality. In the first stage, we pre-enhance a low light image with a conventional Retinex based method. To alleviate these problems, in this paper, we propose a two-stage unsupervised method that decomposes the low light image enhancement into a pre-enhancement and a post-refinement problem. However, most of them suffer from the following problems: 1) the need of pairs of low light and normal light images for training, 2) the poor performance for dark images, 3) the amplification of noise. Recently, deep learning based methods have been proposed to enhance low light images by penalizing the pixel-wise loss of low light and normal light images. Extensive studies demonstrate that our method outperforms the other self-supervised and even unpaired denoising methods by a large margin, without using any additional knowledge, e.g., noise level, regarding the underlying unknown noise.Īs vision based perception methods are usually built on the normal light assumption, there will be a serious safety issue when deploying them into low light environments. We further propose random-replacing refinement, which significantly improves the performance of our AP-BSN without any additional parameters.
To this end, we develop AP-BSN, a state-of-the-art self-supervised denoising method for real-world sRGB images. We systematically demonstrate that the proposed AP can resolve inherent trade-offs caused by specific PD stride factors and make BSN applicable to practical scenarios. We propose an Asymmetric PD (AP) to address this issue, which introduces different PD stride factors for training and inference. However, it is not trivial to integrate PD and BSN directly, which prevents the fully self-supervised denoising model on real-world images. Recently, pixel-shuffle downsampling (PD) has been proposed to remove the spatial correlation of real-world noise. Hence, it is challenging to deal with spatially correlated real-world noise using self-supervised BSN. Nevertheless, they are still bound to synthetic noisy inputs due to less practical assumptions like pixel-wise independent noise. Blind-spot network (BSN) and its variants have made significant advances in self-supervised denoising.
0 kommentar(er)
