{"paper":{"title":"X-Restormer++: 1st Place Solution for the UG2+ CVPR 2026 All-Weather Restoration Challenge","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"X-Restormer++ wins first place in all-weather image restoration by adding adaptive scaling, edge-aware loss, and extra training pairs","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Fengjie Zhu, Leilei Cao, Yingfang Zhu, Youwei Pan","submitted_at":"2026-05-13T09:41:09Z","abstract_excerpt":"In this work, we present our winning solution for the 8th UG2+ Challenge (CVPR 2026) Track 1: Image Restoration under All-weather Conditions. Our method is built upon the strong baseline framework X-Restormer, which effectively captures both channel-wise global dependencies and spatially-local structural information through its dual-attention design (Multi-DConv Head Transposed Attention and Overlapping Cross-Attention). To further boost the restoration performance, we propose several key improvements. First, we integrate the spatially-adaptive input scaling mechanism from Restormer-Plus to dy"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"With these strategies, our proposed method successfully ranks the 1st place in the challenge.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the performance gain is primarily attributable to the three listed changes rather than differences in training schedule, optimizer settings, or undisclosed data filtering that were not ablated.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"X-Restormer++ wins the UG2+ all-weather image restoration challenge by combining adaptive scaling, a gradient-guided edge-aware loss, and expanded training data on top of the X-Restormer baseline.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"X-Restormer++ wins first place in all-weather image restoration by adding adaptive scaling, edge-aware loss, and extra training pairs","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"dc792598c1dbfbe0368f7cd66bf5461579d06fe73a3ba131d27760116243951b"},"source":{"id":"2605.13258","kind":"arxiv","version":1},"verdict":{"id":"1ef25422-d245-4c0f-95d7-d3705014a6f6","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T19:58:22.540307Z","strongest_claim":"With these strategies, our proposed method successfully ranks the 1st place in the challenge.","one_line_summary":"X-Restormer++ wins the UG2+ all-weather image restoration challenge by combining adaptive scaling, a gradient-guided edge-aware loss, and expanded training data on top of the X-Restormer baseline.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the performance gain is primarily attributable to the three listed changes rather than differences in training schedule, optimizer settings, or undisclosed data filtering that were not ablated.","pith_extraction_headline":"X-Restormer++ wins first place in all-weather image restoration by adding adaptive scaling, edge-aware loss, and extra training pairs"},"references":{"count":9,"sample":[{"doi":"","year":2023,"title":"A comparative study of image restoration networks for gen eral backbone network design","work_id":"66c039c6-2486-4d30-98ef-2b3776a085e8","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Activating more pixels in image super-resolution transformer","work_id":"bd178b7c-857e-470e-b847-a70a3806d9a2","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"Weatherbench: A real-world bench- mark dataset for all-in-one adverse weather image restoration","work_id":"45f08c4f-8f16-4a7d-9f9c-765a36fda214","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"Foundir: Unleashing million-scale training data to ad- vance foundation models for image restoration","work_id":"ba18d76b-d4f1-4b9e-8ab9-cf9afe212b89","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Diff- bir: Toward blind image restoration with generative diffusion prior","work_id":"7e0bed32-0c16-415d-a35b-1811e43367d4","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":9,"snapshot_sha256":"6886c2d9e4ba11b1612e65e35e49b2d8b85dc71724ef16b6b9e15c3ecf3dcdcf","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}