Knowledge Commons of Institute of Automation,CAS
Blind2Unblind: Self-Supervised Image Denoising with Visible Blind Spots | |
Wang, Zejin1,2; Liu, Jiazheng1,3; Li, Guoqing1; Han, Hua1,3 | |
2022-06 | |
会议名称 | 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) |
会议日期 | 18-24 June 2022 |
会议地点 | New Orleans, LA, USA |
摘要 | Real noisy-clean pairs on a large scale are costly and difficult to obtain. Meanwhile, supervised denoisers trained on synthetic data perform poorly in practice. Selfsupervised denoisers, which learn only from single noisy images, solve the data collection problem. However, selfsupervised denoising methods, especially blindspot-driven ones, suffer sizable information loss during input or network design. The absence of valuable information dramatically reduces the upper bound of denoising performance. In this paper, we propose a simple yet efficient approach called Blind2Unblind to overcome the information loss in blindspot-driven denoising methods. First, we introduce a global-aware mask mapper that enables global perception and accelerates training. The mask mapper samples all pixels at blind spots on denoised volumes and maps them to the same channel, allowing the loss function to optimize all blind spots at once. Second, we propose a re-visible loss to train the denoising network and make blind spots visible. The denoiser can learn directly from raw noise images without losing information or being trapped in identity mapping. We also theoretically analyze the convergence of the re-visible loss. Extensive experiments on synthetic and real-world datasets demonstrate the superior performance of our approach compared to previous work. Code is available at https://github.com/demonsjin/Blind2Unblind. |
收录类别 | EI |
语种 | 英语 |
是否为代表性论文 | 是 |
七大方向——子方向分类 | 图像视频处理与分析 |
国重实验室规划方向分类 | 视觉信息处理 |
是否有论文关联数据集需要存交 | 否 |
文献类型 | 会议论文 |
条目标识符 | http://ir.ia.ac.cn/handle/173211/51731 |
专题 | 脑图谱与类脑智能实验室_微观重建与智能分析 |
通讯作者 | Han, Hua |
作者单位 | 1.National Laboratory of Pattern Recognition, Chinese Academy of Sciences, Institute of Automation 2.School of Artificial Intelligence, University of Chinese Academy of Sciences 3.School of Future Technology, University of Chinese Academy of Sciences |
第一作者单位 | 中国科学院自动化研究所 |
通讯作者单位 | 中国科学院自动化研究所 |
推荐引用方式 GB/T 7714 | Wang, Zejin,Liu, Jiazheng,Li, Guoqing,et al. Blind2Unblind: Self-Supervised Image Denoising with Visible Blind Spots[C],2022. |
条目包含的文件 | 下载所有文件 | |||||
文件名称/大小 | 文献类型 | 版本类型 | 开放类型 | 使用许可 | ||
Wang_Blind2Unblind_S(7811KB) | 会议论文 | 开放获取 | CC BY-NC-SA | 浏览 下载 |
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。
修改评论