{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2605.30090", "kind": "paper", "url": "https://arxiv.org/abs/2605.30090"}], "authors": ["Jiamin Chen", "Qianben Chen", "Jiawen Zhang", "Yidi Wu", "Yuchen Li", "Xiaokun Zhang", "Wangchunshu Zhou", "Chen Ma"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2605.30090", "source_url": "https://arxiv.org/abs/2605.30090"}, "categories": ["Agents", "General AI", "Vision & 3D"], "description": {"en": "DirectorBench evaluates long-form video generation across 5 dimensions (script, visual, audio, cross-modal, stability) using 80 metadata entries, 7 user profiles, and 40 checkpoint criteria."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2605.30090", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2605.30090", "title": "DirectorBench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["City University of Hong Kong", "ByteDance (China)"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2605.30090", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2605.30090", "source_url": "https://openalex.org/W7162785817"}, "key": "claire-radar:2605.30090", "modality": null, "name": "DirectorBench", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"confidence": "Medium", "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_directorbench_ecf5066a", "origin_source": "arxiv", "recognition_confidence": "0.85", "record_sha256": "837049610774761369fb2699c8deddffe70553634def12cb4a778fc10d1f4e7e", "relation": "introduces", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2605.30090"}, "publisher": null, "released": "2026-05-28", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2605.30090", "source_url": "https://arxiv.org/abs/2605.30090"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2605-30090", "source": "claire_radar", "source_benchmark_id": "2605.30090", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Mixed", "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/2605.30090", "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": "Medium", "construction": "Aggregate Existing", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-08-20T15:22:10.044731Z", "inputHash": "9fc5422221b76b67fcfeb6a02b2898ca0e9f60e6ee23fa1997e0487b04bad29d", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "DirectorBench evaluates long-form video generation across 5 dimensions (script, visual, audio, cross-modal, stability) using 80 metadata entries, 7 user profiles, and 40 checkpoint criteria.", "domainCuration": {"method": "rules-v1", "state": "auto"}, "evidence": {"reasonCodes": ["coined title prefix ending in Bench or Benchmark", "benchmark release stated in one sentence"], "snippet": "We introduce DirectorBench, a personalized multi-agent diagnostic benchmark for long-form video generation."}, "familyId": "bmf_02198acaf891", "firstSeenAt": "2026-08-19", "githubIndex": {"candidateCount": 0, "indexedAt": "2026-09-17T04:40:34.111665+00:00", "repositories": [], "status": "not-found"}, "heat": null, "id": "bm_directorbench_ecf5066a", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": null, "data": null, "hfPaper": "https://huggingface.co/papers/2605.30090", "paper": "https://arxiv.org/abs/2605.30090", "pdf": "https://arxiv.org/pdf/2605.30090", "project": null, "report": "https://arxiv.org/abs/2605.30090"}, "metrics": [], "motivation": "Long-form video generation is rapidly moving from short, single-scene synthesis toward minute-long, multi-shot creation with narrative structure, cinematic control, audio, and cross-modal synchronization.", "name": "DirectorBench", "oneLine": "DirectorBench evaluates long-form video generation across 5 dimensions (script, visual, audio, cross-modal, stability) using 80 metadata entries, 7 user profiles, and 40 checkpoint criteria.", "paperTitle": "DirectorBench: Diagnosing Long-Form Video Generation with Personalized Multi-Agent Evaluation", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2605.30090", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "publishers": [{"name": "arXiv", "organizationType": "community", "role": "benchmark-publisher", "sourceUrl": "https://arxiv.org/abs/2605.30090"}], "ranking": {}, "readiness": "Paper only", "recognitionConfidence": 0.85, "relation": "introduces", "releasedAt": "2026-05-28", "source": {"authors": ["Jiamin Chen", "Qianben Chen", "Jiawen Zhang", "Yidi Wu", "Yuchen Li", "Xiaokun Zhang", "Wangchunshu Zhou", "Chen Ma"], "categories": ["cs.CL", "cs.CV"], "id": "2605.30090", "title": "DirectorBench: Diagnosing Long-Form Video Generation with Personalized Multi-Agent Evaluation", "type": "arxiv", "url": "https://arxiv.org/abs/2605.30090"}, "sourceUpdatedAt": "2026-05-28T15:35:34Z", "topics": ["Agents"], "venueAttempts": [], "whyItMatters": "Current video benchmarks focus on short clips and aggregate scores, missing workflow failures and user preferences. DirectorBench provides diagnostic, profile-aware evaluation to guide improvements in long-form video generation."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
