{"record": {"aliases": ["TAPVid-MV: A Benchmark for Tracking Any Point in 3D Across Multiple Views"], "artifacts": [{"id": "arxiv:2609.01899", "kind": "paper", "url": "https://arxiv.org/abs/2609.01899"}], "authors": ["Skanda Koppula", "Frano Rajic", "Abdullah Faiz Ur Rahman", "Yi Yang", "Ignacio Rocco", "Jeet Thakwani", "Rishabh Kabra", "Andrew Zisserman", "Joao Carreira", "Siyu Tang", "Carl Doersch", "Gabriel Brostow"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2609.01899", "source_url": "https://arxiv.org/abs/2609.01899"}, "categories": ["General AI", "Geometric reasoning", "Vision & 3D", "cs.CV"], "description": {"en": "Evaluates long-term 3D point tracking across multiple synchronized views with camera motion using 284 sequences and over 100K point tracks."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2609.01899", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2609.01899", "title": "TAPVid-MV"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2609.01899", "modality": null, "name": "TAPVid-MV", "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": "true", "origin_record_id": "bm_tapvid-mv-a-benchmark-for-tracking-any-poi_a0341f90", "origin_source": "arxiv", "recognition_confidence": "0.6", "record_sha256": "5f3d47b3fcb6a551b04f3970340646179c94b0ae83179926fd25e36ca9f40d9e", "relation": "introduces", "review_model": "deepseek-v4-pro", "review_state": "ai-reviewed", "reviewed_at": "2026-09-03T06:06:59.591745Z", "source_url": "https://arxiv.org/abs/2609.01899"}, "publisher": null, "released": "2026-09-01", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2609.01899", "source_url": "https://arxiv.org/abs/2609.01899"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2609-01899", "source": "claire_radar", "source_benchmark_id": "2609.01899", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": ["TAPVid-MV: A Benchmark for Tracking Any Point in 3D Across Multiple Views"], "annotation": "Machine Generated", "applicationDomains": ["General AI"], "area": "Vision & 3D", "attention": {"asOf": "2026-09-03", "attemptedAt": "2026-09-03T01:40:48.194060Z", "githubRepo": null, "githubScope": null, "githubStars": null, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": null, "hfPaperUrl": "https://huggingface.co/papers/2609.01899", "observedAt": "2026-09-02T06:30:32.214812Z", "signalStatus": {"githubStars": {"state": "not_applicable"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"checkedAt": "2026-09-03T01:40:48.194060Z", "state": "unavailable"}}, "status": "unavailable"}, "attentionForecast": {"confidence": "Medium", "horizon": "7d", "reason": "The benchmark tackles a core problem in 3D vision with large-scale annotations and broad applicability, likely to draw attention from the tracking and reconstruction communities.", "score": 60}, "capabilities": ["Geometric reasoning"], "confidence": "Medium", "construction": "Original Synthetic", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-03T06:06:59.591745Z", "inputHash": "7e03667503118791a8c3296088051831cd62f2c2df919f9660223dfb344a0ff3", "model": "deepseek-v4-pro"}, "curation": {"canonicalNameEvidence": "TAPVid-MV: A Benchmark for Tracking Any Point in 3D Across Multiple Views", "canonicalNameSource": "paper_title", "decisionReason": "Comprehensive named benchmark with released annotations and clear public path through paper and dataset.", "model": "deepseek-v4-pro", "reviewedAt": "2026-09-03T06:06:59.591745Z", "state": "ai-reviewed"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "Evaluates long-term 3D point tracking across multiple synchronized views with camera motion using 284 sequences and over 100K point tracks.", "displayEligible": true, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "score_submission", "evidence": {"reasonCodes": ["benchmark term in abstract", "benchmark release stated in one sentence", "evaluation protocol evidence"], "snippet": "We introduce TAPVid-MV (Tracking Any Point in Video across Multiple Views), the first benchmark for this setting."}, "familyId": "bmf_64eae251580e", "firstSeenAt": "2026-09-03", "heat": null, "id": "bm_tapvid-mv-a-benchmark-for-tracking-any-poi_a0341f90", "indexedAt": "2026-09-03T00:55:28.123857Z", "industrySectors": [], "links": {"code": null, "data": null, "paper": "https://arxiv.org/abs/2609.01899", "pdf": "https://arxiv.org/pdf/2609.01899", "project": null, "report": "https://arxiv.org/abs/2609.01899"}, "metrics": [], "motivation": "Multi-camera systems are increasingly practical for robotics, AR/VR, and autonomous driving because complementary views reduce depth ambiguity and preserve visibility under occlusion.", "name": "TAPVid-MV", "oneLine": "Evaluates long-term 3D point tracking across multiple synchronized views with camera motion using 284 sequences and over 100K point tracks.", "paperTitle": "TAPVid-MV: A Benchmark for Tracking Any Point in 3D Across Multiple Views", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2609.01899", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-09-03T00:55:28.123857Z"}, "publications": [], "ranking": {"30d": {"confidence": "Low", "coverage": 0.0, "growth": {"components": {"githubStars": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfDatasetDownloads": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfPaperUpvotes": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "elapsedDays": null, "historyDate": null, "rank": null, "score": null, "windowDays": 30}, "level": {"components": {"githubStars": {"normalization": "log-max", "normalized": null, "value": null}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "rank": null, "score": null}, "method": "attention-ranking-v13", "rank": null, "score": null}, "90d": {"confidence": "Low", "coverage": 0.0, "growth": {"components": {"githubStars": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfDatasetDownloads": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfPaperUpvotes": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "elapsedDays": null, "historyDate": null, "rank": null, "score": null, "windowDays": 90}, "level": {"components": {"githubStars": {"normalization": "log-max", "normalized": null, "value": null}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "rank": null, "score": null}, "method": "attention-ranking-v13", "rank": null, "score": null}}, "readiness": "Paper only", "recognitionConfidence": 0.6, "relation": "introduces", "releasedAt": "2026-09-01", "source": {"authors": ["Skanda Koppula", "Frano Rajic", "Abdullah Faiz Ur Rahman", "Yi Yang", "Ignacio Rocco", "Jeet Thakwani", "Rishabh Kabra", "Andrew Zisserman", "Joao Carreira", "Siyu Tang", "Carl Doersch", "Gabriel Brostow"], "categories": ["cs.CV"], "id": "2609.01899", "title": "TAPVid-MV: A Benchmark for Tracking Any Point in 3D Across Multiple Views", "type": "arxiv", "url": "https://arxiv.org/abs/2609.01899"}, "sourceUpdatedAt": "2026-09-01T21:58:51Z", "topics": ["cs.CV"], "venueAttempts": [], "whyItMatters": "Addresses a gap in multi-view tracking benchmarks, distinguishing geometry recovery from correspondence errors and enabling joint analysis."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
