{"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/jafar-jack-up-any-feature-at-any-resolution-1","title":"JAFAR: Jack up Any Feature at Any Resolution","arxiv_id":"2506.11136","date":"2025-06-10","proceeding":null,"authors":["Paul Couairon","Loick Chambon","Louis Serrano","Jean-Emmanuel Haugeard","Matthieu Cord","Nicolas Thome"],"abstract":"Foundation Vision Encoders have become essential for a wide range of dense vision tasks. However, their low-resolution spatial feature outputs necessitate feature upsampling to produce the high-resolution modalities required for downstream tasks. In this work, we introduce JAFAR, a lightweight and flexible feature upsampler that enhances the spatial resolution of visual features from any Foundation Vision Encoder to an arbitrary target resolution. JAFAR employs an attention-based module designed to promote semantic alignment between high-resolution queries, derived from low-level image features, and semantically enriched low-resolution keys, using Spatial Feature Transform (SFT) modulation. Notably, despite the absence of high-resolution supervision, we demonstrate that learning at low upsampling ratios and resolutions generalizes remarkably well to significantly higher output scales. Extensive experiments show that JAFAR effectively recovers fine-grained spatial details and consistently outperforms existing feature upsampling methods across a diverse set of downstream tasks. Project page at https://jafar-upsampler.github.io","url_abs":"https://arxiv.org/abs/2506.11136v1","url_pdf":"https://arxiv.org/pdf/2506.11136v1.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":"jafar-jack-up-any-feature-at-any-resolution-1","repo_url":"https://github.com/PaulCouairon/JAFAR","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"NOASSERTION"}}],"tasks":[{"task_slug":"feature-upsampling","task_name":"Feature Upsampling"}],"methods":[{"method_slug":"set","method_name":"SET"},{"method_slug":"spatial-feature-transform","method_name":"Spatial Feature Transform"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/feature-upsampling-on-imagenet","task":"Feature Upsampling","dataset":"ImageNet","model":"JAFAR","rank_in_archive_order":1,"of":8,"metrics":{"ADCC":"73.3","Average Drop":"17.4","Average Increase":"30.9"},"uses_additional_data":false}],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=2506.11136","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}