{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2606.02535", "kind": "paper", "url": "https://arxiv.org/abs/2606.02535"}], "authors": ["Lu Liu", "Huiyu Duan", "Chenxin Zhu", "Jintong Lu", "Haoyun Jiang", "Liu Yang", "Qiang Hu", "Guangtao Zhai", "Xiaoyun Zhang"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2606.02535", "source_url": "https://arxiv.org/abs/2606.02535"}, "categories": ["General AI", "Vision & 3D", "cs.CV"], "description": {"en": "LL-Bench evaluates large-scale generative models on 16 low-level vision tasks using 2,469 real-world degraded images, with human preference and quality score annotations for model outputs."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2606.02535", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2606.02535", "title": "LL-Bench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["Shanghai Jiao Tong University"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2606.02535", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2606.02535", "source_url": "https://openalex.org/W7163212961"}, "key": "claire-radar:2606.02535", "modality": null, "name": "LL-Bench", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"confidence": "High", "crawl_bundle": "claire_radar_2026-10-14", "crawled_at": "2026-10-14T09:44:51Z", "data_status": "primary-source-indexed", "display_eligible": "unknown", "origin_record_id": "bm_ll-bench_df7573ac", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "0fcbdf1f8996d45edff8a8c23f1bfa54b6e20312a3280bdc4774fd924664e1df", "relation": "introduces", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2606.02535"}, "publisher": null, "released": "2026-06-01", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2606.02535", "source_url": "https://arxiv.org/abs/2606.02535"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2606-02535", "source": "claire_radar", "source_benchmark_id": "2606.02535", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Vision & 3D", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": null, "githubScope": null, "githubStars": null, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": null, "hfPaperUrl": "https://huggingface.co/papers/2606.02535", "observedAt": "2026-10-04T03:02:42.754940Z", "signalStatus": {"githubStars": {"state": "not_applicable"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"checkedAt": "2026-10-04T15:31:48.030916Z", "state": "unavailable"}}, "status": "unavailable"}, "capabilities": [], "confidence": "High", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-08-20T15:22:10.044731Z", "inputHash": "7ba0d380642efecde1cd833658bd564758243f39f900f18ad4067939ea0eda2d", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "LL-Bench evaluates large-scale generative models on 16 low-level vision tasks using 2,469 real-world degraded images, with human preference and quality score annotations for model outputs.", "domainCuration": {"method": "rules-v1", "state": "auto"}, "evidence": {"reasonCodes": ["coined title prefix ending in Bench or Benchmark", "benchmark release stated in one sentence", "evaluation protocol evidence"], "snippet": "To address this gap, we introduce \\textbf{LL-Bench}, a comprehensive \\textbf{Benchmark} for evaluating the capabilities of large-scale generative models on \\textbf{L}ow-\\textbf{L}evel vision tasks."}, "familyId": "bmf_e383e91589a5", "firstSeenAt": "2026-08-19", "githubIndex": {"candidateCount": 0, "indexedAt": "2026-09-17T04:40:34.111665+00:00", "repositories": [], "status": "not-found"}, "heat": null, "id": "bm_ll-bench_df7573ac", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": null, "data": null, "hfPaper": "https://huggingface.co/papers/2606.02535", "paper": "https://arxiv.org/abs/2606.02535", "pdf": "https://arxiv.org/pdf/2606.02535", "project": null, "report": "https://arxiv.org/abs/2606.02535"}, "metrics": [], "motivation": "Large-scale generative models have demonstrated remarkable capabilities across image generation and editing tasks.", "name": "LL-Bench", "oneLine": "LL-Bench evaluates large-scale generative models on 16 low-level vision tasks using 2,469 real-world degraded images, with human preference and quality score annotations for model outputs.", "paperTitle": "LL-Bench: Rethinking Low-Level Vision Evaluation in the Era of Large-Scale Generative Models", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2606.02535", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "ranking": {}, "readiness": "Paper only", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-06-01", "source": {"authors": ["Lu Liu", "Huiyu Duan", "Chenxin Zhu", "Jintong Lu", "Haoyun Jiang", "Liu Yang", "Qiang Hu", "Guangtao Zhai", "Xiaoyun Zhang"], "categories": ["cs.CV"], "id": "2606.02535", "title": "LL-Bench: Rethinking Low-Level Vision Evaluation in the Era of Large-Scale Generative Models", "type": "arxiv", "url": "https://arxiv.org/abs/2606.02535"}, "sourceUpdatedAt": "2026-06-01T17:40:09Z", "topics": ["cs.CV"], "venueAttempts": [], "whyItMatters": "Existing low-level vision benchmarks often focus on conventional models and lack alignment with human perception. LL-Bench provides a standardized evaluation suite to compare generative and restoration models on pixel-level tasks, supporting quality assessment and model selection."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
