{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/nonblind-image-deconvolution-via-leveraging","title":"Nonblind image deconvolution via leveraging model uncertainty in an untrained deep neural network","arxiv_id":null,"date":"2022-05-18","proceeding":"International Journal of Computer Vision 2022 5","authors":["Mingqin Chen; Yuhui Quan; Tongyao Pang; Hui Ji"],"abstract":"Nonblind image deconvolution (NID) is about restoring the latent image with sharp details from\r\na noisy blurred one using a known blur kernel. This paper presents a dataset-free deep learning\r\napproach for NID using untrained deep neural networks (DNNs), which does not require any external\r\ntraining data with ground-truth images. Based on a spatially-adaptive dropout scheme, the proposed\r\napproach learns a DNN with model uncertainty from the input blurred image, and the deconvolution result is obtained by aggregating the multiple predictions from the learned dropout DNN.\r\nIt is shown that the solution approximates a minimum-mean-squared-error estimator in Bayesian\r\ninference. In addition, a self-supervised loss function for training is presented to efficiently handle\r\nthe noise in blurred images. Extensive experiments show that the proposed approach not only performs noticeably better than existing non-learning-based methods and unsupervised learning-based\r\nmethods, but also performs competitively against recent supervised learning-based methods.","url_abs":"https://link.springer.com/article/10.1007/s11263-022-01621-9","url_pdf":"https://link.springer.com/article/10.1007/s11263-022-01621-9","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"nonblind-image-deconvolution-via-leveraging","repo_url":"https://github.com/scut-mingqinchen/Model_Uncertainty_NID","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"bayesian-inference","task_name":"Bayesian Inference"},{"task_slug":"image-deconvolution","task_name":"Image Deconvolution"}],"methods":[{"method_slug":"dropout","method_name":"Dropout"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}