{"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/generative-collaborative-networks-for-single","title":"Generative Collaborative Networks for Single Image Super-Resolution","arxiv_id":"1902.10467","date":"2019-02-27","proceeding":null,"authors":["Mohamed El Amine Seddik","Mohamed Tamaazousti","John Lin"],"abstract":"A common issue of deep neural networks-based methods for the problem of\nSingle Image Super-Resolution (SISR), is the recovery of finer texture details\nwhen super-resolving at large upscaling factors. This issue is particularly\nrelated to the choice of the objective loss function. In particular, recent\nworks proposed the use of a VGG loss which consists in minimizing the error\nbetween the generated high resolution images and ground-truth in the feature\nspace of a Convolutional Neural Network (VGG19), pre-trained on the very\n\"large\" ImageNet dataset. When considering the problem of super-resolving\nimages with a distribution \"far\" from the ImageNet images distribution\n(\\textit{e.g.,} satellite images), their proposed \\textit{fixed} VGG loss is no\nlonger relevant. In this paper, we present a general framework named\n\\textit{Generative Collaborative Networks} (GCN), where the idea consists in\noptimizing the \\textit{generator} (the mapping of interest) in the feature\nspace of a \\textit{features extractor} network. The two networks (generator and\nextractor) are \\textit{collaborative} in the sense that the latter \"helps\" the\nformer, by constructing discriminative and relevant features (not necessarily\n\\textit{fixed} and possibly learned \\textit{mutually} with the generator). We\nevaluate the GCN framework in the context of SISR, and we show that it results\nin a method that is adapted to super-resolution domains that are \"far\" from the\nImageNet domain.","url_abs":"http://arxiv.org/abs/1902.10467v2","url_pdf":"http://arxiv.org/pdf/1902.10467v2.pdf","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":"generative-collaborative-networks-for-single","repo_url":"https://github.com/melaseddik/GCN","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[{"task_slug":"image-super-resolution","task_name":"Image Super-Resolution"},{"task_slug":"super-resolution","task_name":"Super-Resolution"}],"methods":[{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"dense-connections","method_name":"Dense Connections"},{"method_slug":"dropout","method_name":"Dropout"},{"method_slug":"gcn","method_name":"GCN"},{"method_slug":"max-pooling","method_name":"Max Pooling"},{"method_slug":"relu","method_name":"ReLU"},{"method_slug":"softmax","method_name":"Softmax"},{"method_slug":"vgg-loss","method_name":"VGG Loss"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1902.10467","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}