{"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/on-efficient-transformer-and-image-pre","title":"On Efficient Transformer-Based Image Pre-training for Low-Level Vision","arxiv_id":"2112.10175","date":"2021-12-19","proceeding":null,"authors":["Wenbo Li","Xin Lu","Shengju Qian","Jiangbo Lu","Xiangyu Zhang","Jiaya Jia"],"abstract":"Pre-training has marked numerous state of the arts in high-level computer vision, while few attempts have ever been made to investigate how pre-training acts in image processing systems. In this paper, we tailor transformer-based pre-training regimes that boost various low-level tasks. To comprehensively diagnose the influence of pre-training, we design a whole set of principled evaluation tools that uncover its effects on internal representations. The observations demonstrate that pre-training plays strikingly different roles in low-level tasks. For example, pre-training introduces more local information to higher layers in super-resolution (SR), yielding significant performance gains, while pre-training hardly affects internal feature representations in denoising, resulting in limited gains. Further, we explore different methods of pre-training, revealing that multi-related-task pre-training is more effective and data-efficient than other alternatives. Finally, we extend our study to varying data scales and model sizes, as well as comparisons between transformers and CNNs-based architectures. Based on the study, we successfully develop state-of-the-art models for multiple low-level tasks. Code is released at https://github.com/fenglinglwb/EDT.","url_abs":"https://arxiv.org/abs/2112.10175v2","url_pdf":"https://arxiv.org/pdf/2112.10175v2.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":"on-efficient-transformer-and-image-pre","repo_url":"https://github.com/fenglinglwb/edt","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"denoising","task_name":"Denoising"},{"task_slug":"image-super-resolution","task_name":"Image Super-Resolution"},{"task_slug":"super-resolution","task_name":"Super-Resolution"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/image-super-resolution-on-set5-2x-upscaling","task":"Image Super-Resolution","dataset":"Set5 - 2x upscaling","model":"EDT-B","rank_in_archive_order":11,"of":41,"metrics":{"PSNR":"38.63","SSIM":"0.9632"},"uses_additional_data":true},{"leaderboard":"/sota/image-super-resolution-on-set5-3x-upscaling","task":"Image Super-Resolution","dataset":"Set5 - 3x upscaling","model":"EDT-B","rank_in_archive_order":11,"of":32,"metrics":{"PSNR":"35.13","SSIM":"0.9328"},"uses_additional_data":true}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2112.10175","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}