{"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/rtformer-efficient-design-for-real-time","title":"RTFormer: Efficient Design for Real-Time Semantic Segmentation with Transformer","arxiv_id":"2210.07124","date":"2022-10-13","proceeding":null,"authors":["Jian Wang","Chenhui Gou","Qiman Wu","Haocheng Feng","Junyu Han","Errui Ding","Jingdong Wang"],"abstract":"Recently, transformer-based networks have shown impressive results in semantic segmentation. Yet for real-time semantic segmentation, pure CNN-based approaches still dominate in this field, due to the time-consuming computation mechanism of transformer. We propose RTFormer, an efficient dual-resolution transformer for real-time semantic segmenation, which achieves better trade-off between performance and efficiency than CNN-based models. To achieve high inference efficiency on GPU-like devices, our RTFormer leverages GPU-Friendly Attention with linear complexity and discards the multi-head mechanism. Besides, we find that cross-resolution attention is more efficient to gather global context information for high-resolution branch by spreading the high level knowledge learned from low-resolution branch. Extensive experiments on mainstream benchmarks demonstrate the effectiveness of our proposed RTFormer, it achieves state-of-the-art on Cityscapes, CamVid and COCOStuff, and shows promising results on ADE20K. Code is available at PaddleSeg: https://github.com/PaddlePaddle/PaddleSeg.","url_abs":"https://arxiv.org/abs/2210.07124v1","url_pdf":"https://arxiv.org/pdf/2210.07124v1.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":"rtformer-efficient-design-for-real-time","repo_url":"https://github.com/PaddlePaddle/PaddleSeg","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"paddle","reach":null}],"tasks":[{"task_slug":null,"task_name":"GPU"},{"task_slug":"real-time-semantic-segmentation","task_name":"Real-Time Semantic Segmentation"},{"task_slug":"segmentation","task_name":"Segmentation"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/real-time-semantic-segmentation-on-camvid","task":"Real-Time Semantic Segmentation","dataset":"CamVid","model":"RTFormer-Slim","rank_in_archive_order":2,"of":29,"metrics":{"Frame (fps)":"190.7(2080Ti)","mIoU":"81.4"},"uses_additional_data":false},{"leaderboard":"/sota/real-time-semantic-segmentation-on-cityscapes-1","task":"Real-Time Semantic Segmentation","dataset":"Cityscapes val","model":"RTFormer-B","rank_in_archive_order":6,"of":24,"metrics":{"Frame (fps)":"50.2","mIoU":"79.3%"},"uses_additional_data":false},{"leaderboard":"/sota/real-time-semantic-segmentation-on-cityscapes-1","task":"Real-Time Semantic Segmentation","dataset":"Cityscapes val","model":"RTFormer-S","rank_in_archive_order":13,"of":24,"metrics":{"Frame (fps)":"89.6","mIoU":"76.3%"},"uses_additional_data":false},{"leaderboard":"/sota/semantic-segmentation-on-camvid","task":"Semantic Segmentation","dataset":"CamVid","model":"RTFormer-Base","rank_in_archive_order":4,"of":21,"metrics":{"Mean IoU":"82.5"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2210.07124","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}