{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2606.15320", "kind": "paper", "url": "https://arxiv.org/abs/2606.15320"}], "authors": ["Yuanhao Zou", "Arthad Kulkarni", "Lucas Tonanez", "Lincoln Spencer", "Guangyu Sun", "Tianxingjian Ding", "Andong Deng", "Yi Li", "Shuangjun Liu", "Yuan Li", "Dashan Gao", "Ning Bi", "Taotao Jing", "Shuai Zhang", "Chen Chen"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2606.15320", "source_url": "https://arxiv.org/abs/2606.15320"}, "categories": ["General AI", "Vision & 3D", "cs.CV"], "description": {"en": "CoMET-Bench is a benchmark for conditional multi-event temporal grounding in long-form video, with 2,789 queries over 600 videos and a unified evaluation protocol including counting, grounding, and negative-query recognition."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2606.15320", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2606.15320", "title": "CoMET-Bench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["University of Central Florida", "Qualcomm (United States)"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2606.15320", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2606.15320", "source_url": "https://openalex.org/W7164915235"}, "key": "claire-radar:2606.15320", "modality": null, "name": "CoMET-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_comet-bench_355dabd3", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "40155ea7b26dc68da2976cd70690c50e42fd42e69e9f5d85827ea60b8e50368b", "relation": "introduces", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2606.15320"}, "publisher": null, "released": "2026-06-13", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2606.15320", "source_url": "https://arxiv.org/abs/2606.15320"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2606-15320", "source": "claire_radar", "source_benchmark_id": "2606.15320", "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/2606.15320", "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": "Aggregate Existing", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-08-20T15:22:10.044731Z", "inputHash": "58c5389c132002ab78491d9e7721fe06ff02127749af1a073657480c0b943e1d", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "CoMET-Bench is a benchmark for conditional multi-event temporal grounding in long-form video, with 2,789 queries over 600 videos and a unified evaluation protocol including counting, grounding, and negative-query recognition.", "domainCuration": {"method": "rules-v1", "state": "auto"}, "evidence": {"reasonCodes": ["exact named benchmark artifact released in abstract", "benchmark release stated in one sentence", "evaluation protocol evidence"], "snippet": "We introduce CoMET-Bench for Conditional Multi-Event Temporal Grounding in long-form video, comprising 2789 queries over 600 videos averaging 33.8 minutes across five real-world domains, with each query composed from 4 temporal conditions, 3 spatial conditions, and a dedicated negative-query subset."}, "familyId": "bmf_2bde53728ad5", "firstSeenAt": "2026-08-19", "githubIndex": {"candidateCount": 0, "indexedAt": "2026-09-17T04:40:34.111665+00:00", "repositories": [], "status": "not-found"}, "heat": null, "id": "bm_comet-bench_355dabd3", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": null, "data": null, "hfPaper": "https://huggingface.co/papers/2606.15320", "paper": "https://arxiv.org/abs/2606.15320", "pdf": "https://arxiv.org/pdf/2606.15320", "project": null, "report": "https://arxiv.org/abs/2606.15320"}, "metrics": [], "motivation": "Multimodal large language models have made rapid progress in video temporal grounding, yet real-world applications routinely require localizing every event that satisfies compositional temporal and spatial conditions.", "name": "CoMET-Bench", "oneLine": "CoMET-Bench is a benchmark for conditional multi-event temporal grounding in long-form video, with 2,789 queries over 600 videos and a unified evaluation protocol including counting, grounding, and negative-query recognition.", "paperTitle": "Conditional Multi-Event Temporal Grounding in Long-Form Video", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2606.15320", "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": "2026-06-13", "source": {"authors": ["Yuanhao Zou", "Arthad Kulkarni", "Lucas Tonanez", "Lincoln Spencer", "Guangyu Sun", "Tianxingjian Ding", "Andong Deng", "Yi Li", "Shuangjun Liu", "Yuan Li", "Dashan Gao", "Ning Bi", "Taotao Jing", "Shuai Zhang", "Chen Chen"], "categories": ["cs.CV"], "id": "2606.15320", "title": "Conditional Multi-Event Temporal Grounding in Long-Form Video", "type": "arxiv", "url": "https://arxiv.org/abs/2606.15320"}, "sourceUpdatedAt": "2026-06-13T14:27:58Z", "topics": ["cs.CV"], "venueAttempts": [], "whyItMatters": "Real-world video grounding requires localizing every event satisfying compositional conditions, which existing benchmarks do not jointly handle. This benchmark introduces Rejection-F1 to prevent trivial gaming and exposes gaps in current methods."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
