{"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/local-class-specific-and-global-image-level","title":"Local Class-Specific and Global Image-Level Generative Adversarial Networks for Semantic-Guided Scene Generation","arxiv_id":"1912.12215","date":"2019-12-27","proceeding":"CVPR 2020 6","authors":["Hao Tang","Dan Xu","Yan Yan","Philip H. S. Torr","Nicu Sebe"],"abstract":"In this paper, we address the task of semantic-guided scene generation. One open challenge in scene generation is the difficulty of the generation of small objects and detailed local texture, which has been widely observed in global image-level generation methods. To tackle this issue, in this work we consider learning the scene generation in a local context, and correspondingly design a local class-specific generative network with semantic maps as a guidance, which separately constructs and learns sub-generators concentrating on the generation of different classes, and is able to provide more scene details. To learn more discriminative class-specific feature representations for the local generation, a novel classification module is also proposed. To combine the advantage of both the global image-level and the local class-specific generation, a joint generation network is designed with an attention fusion module and a dual-discriminator structure embedded. Extensive experiments on two scene image generation tasks show superior generation performance of the proposed model. The state-of-the-art results are established by large margins on both tasks and on challenging public benchmarks. The source code and trained models are available at https://github.com/Ha0Tang/LGGAN.","url_abs":"https://arxiv.org/abs/1912.12215v3","url_pdf":"https://arxiv.org/pdf/1912.12215v3.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":"local-class-specific-and-global-image-level","repo_url":"https://github.com/Ha0Tang/LGGAN","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"NOASSERTION"}},{"paper_slug":"local-class-specific-and-global-image-level","repo_url":"https://github.com/Ha0Tang/LocalGlobalGAN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"NOASSERTION"}}],"tasks":[{"task_slug":"image-generation","task_name":"Image Generation"},{"task_slug":"scene-generation","task_name":"Scene Generation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/cross-view-image-to-image-translation-on-2","task":"Cross-View Image-to-Image Translation","dataset":"Dayton (256×256) - aerial-to-ground","model":"LGGAN","rank_in_archive_order":2,"of":6,"metrics":{"KL":"2.18","PSNR":"22.9949","SD":"19.6145","SSIM":"0.5457"},"uses_additional_data":false},{"leaderboard":"/sota/cross-view-image-to-image-translation-on-4","task":"Cross-View Image-to-Image Translation","dataset":"cvusa","model":"LGGAN","rank_in_archive_order":3,"of":7,"metrics":{"KL":"2.55","PSNR":"22.5766","SD":"19.744","SSIM":"0.5238"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/1912.12215","atlas_url":"https://app.syntology.ai/?focus=1912.12215","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}