{"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/unlocking-pre-trained-image-backbones-for","title":"Unlocking Pre-trained Image Backbones for Semantic Image Synthesis","arxiv_id":"2312.13314","date":"2023-12-20","proceeding":"CVPR 2024 1","authors":["Tariq Berrada","Jakob Verbeek","Camille Couprie","Karteek Alahari"],"abstract":"Semantic image synthesis, i.e., generating images from user-provided semantic label maps, is an important conditional image generation task as it allows to control both the content as well as the spatial layout of generated images. Although diffusion models have pushed the state of the art in generative image modeling, the iterative nature of their inference process makes them computationally demanding. Other approaches such as GANs are more efficient as they only need a single feed-forward pass for generation, but the image quality tends to suffer on large and diverse datasets. In this work, we propose a new class of GAN discriminators for semantic image synthesis that generates highly realistic images by exploiting feature backbone networks pre-trained for tasks such as image classification. We also introduce a new generator architecture with better context modeling and using cross-attention to inject noise into latent variables, leading to more diverse generated images. Our model, which we dub DP-SIMS, achieves state-of-the-art results in terms of image quality and consistency with the input label maps on ADE-20K, COCO-Stuff, and Cityscapes, surpassing recent diffusion models while requiring two orders of magnitude less compute for inference.","url_abs":"https://arxiv.org/abs/2312.13314v2","url_pdf":"https://arxiv.org/pdf/2312.13314v2.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":[],"tasks":[{"task_slug":"conditional-image-generation","task_name":"Conditional Image Generation"},{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"image-generation","task_name":"Image Generation"},{"task_slug":"image-to-image-translation","task_name":"Image-to-Image Translation"},{"task_slug":"image-classification","task_name":"image-classification"}],"methods":[{"method_slug":"diffusion","method_name":"Diffusion"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/image-to-image-translation-on-ade20k-labels","task":"Image-to-Image Translation","dataset":"ADE20K Labels-to-Photos","model":"DP-SIMS (ConvNext-L)","rank_in_archive_order":1,"of":16,"metrics":{"FID":"22.7","mIoU":"54.3"},"uses_additional_data":false},{"leaderboard":"/sota/image-to-image-translation-on-coco-stuff","task":"Image-to-Image Translation","dataset":"COCO-Stuff Labels-to-Photos","model":"DP-SIMS (ConvNext-XL)","rank_in_archive_order":1,"of":15,"metrics":{"FID":"13.3"},"uses_additional_data":false},{"leaderboard":"/sota/image-to-image-translation-on-coco-stuff","task":"Image-to-Image Translation","dataset":"COCO-Stuff Labels-to-Photos","model":"DP-SIMS (ConvNext-L)","rank_in_archive_order":2,"of":15,"metrics":{"FID":"13.6"},"uses_additional_data":false},{"leaderboard":"/sota/image-to-image-translation-on-cityscapes","task":"Image-to-Image Translation","dataset":"Cityscapes Labels-to-Photo","model":"DP-SIMS (ConvNext-L)","rank_in_archive_order":1,"of":21,"metrics":{"FID":"38.2","mIoU":"76.3"},"uses_additional_data":false}],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}