{"record": {"aliases": ["Re-engineering SORT-based algorithms for low-cost small object tracking from omnidirectional footage"], "artifacts": [{"id": "arxiv:2609.07547", "kind": "paper", "url": "https://arxiv.org/abs/2609.07547"}, {"id": "gh:xin-shu/omnisort", "kind": "repo", "url": "https://github.com/Xin-Shu/OmniSORT.git"}], "authors": ["Xin Shu", "Meegan Gower", "Yvonne Buckley", "Anil Kokaram"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2609.07547", "source_url": "https://arxiv.org/abs/2609.07547"}, "categories": ["Autonomous Driving", "Geometric reasoning", "Vision & 3D", "cs.CV"], "description": {"en": "OmniSmall evaluates multi-object tracking of flying animals in omnidirectional video, focusing on seam-aware motion and association for small, fast-moving targets."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2609.07547", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2609.07547", "title": "OmniSmall"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["Trinity College Dublin"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2609.07547", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2609.07547", "source_url": "https://openalex.org/W7212042513"}, "key": "claire-radar:2609.07547", "modality": null, "name": "OmniSmall", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"admission_policy_version": "2026-09-09.1", "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_re-engineering-sort-based-algorithms-for-l_55c64c2b", "origin_source": "arxiv", "recognition_confidence": "0.55", "record_sha256": "0007dffbf217ccf30f8400ee87c0bf8dcb63c909adc84860613c29c207ab7600", "relation": "introduces", "review_model": "deepseek-v4-pro", "review_state": "ai-reviewed", "reviewed_at": "2026-09-10T00:52:28.558291Z", "source_url": "https://arxiv.org/abs/2609.07547"}, "publisher": null, "released": "2026-09-07", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2609.07547", "source_url": "https://arxiv.org/abs/2609.07547"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2609-07547", "source": "claire_radar", "source_benchmark_id": "2609.07547", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": ["Re-engineering SORT-based algorithms for low-cost small object tracking from omnidirectional footage"], "annotation": "Unknown", "applicationDomains": ["Autonomous Driving"], "area": "Vision & 3D", "attention": {"asOf": "2026-09-09", "attemptedAt": "2026-09-09T06:48:27.347268Z", "githubRepo": "https://github.com/Xin-Shu/OmniSORT", "githubScope": "benchmark_repo", "githubStars": 1, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": null, "hfPaperUrl": "https://huggingface.co/papers/2609.07547", "observedAt": "2026-09-09T06:48:27.347268Z", "signalStatus": {"githubStars": {"observedAt": "2026-09-09T06:48:27.347268Z", "state": "fresh"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"checkedAt": "2026-09-09T06:48:27.347268Z", "state": "unavailable"}}, "status": "fresh"}, "attentionForecast": {"confidence": "Medium", "horizon": "7d", "reason": "The benchmark targets a niche but increasingly relevant omnidirectional tracking domain, with clear improvements and public artifacts, likely attracting moderate attention from MOT and wildlife monitoring communities.", "score": 65}, "capabilities": ["Geometric reasoning"], "confidence": "Medium", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-10T00:52:28.558291Z", "inputHash": "1f830d5cb04479c3697e9217956449bb524e17c45bfee480299668f8043b2810", "model": "deepseek-v4-pro"}, "curation": {"admissionPolicyVersion": "2026-09-09.1", "canonicalNameEvidence": "including (i) a Seam-Aware Motion Model that keeps the Kalman state continuous across the seam, (ii) a composite seam-aware association cost that pairs a wrapped Euclidean term with GIoU, and (iii) OmniSmall, a new benchmark of omnidirectional wildlife footage.", "canonicalNameSource": "abstract", "decisionReason": "Released benchmark with dataset and code, explicit scoring metrics, and a public repository for inspection and reuse.", "evaluationEvidence": {"reuse": "Dataset and source code are available at a linked GitHub repository, providing a public path for other teams to run or inspect the benchmark.", "scoring": "Uses HOTA, MOTA, and IDF1 metrics, as indicated by improvements over OCSORT by +8.51 HOTA, +9.41 MOTA, and +10.17 IDF1 with ground-truth detections.", "task": "Multi-object tracking of flying animals in omnidirectional wildlife footage, evaluating small and fast-moving targets across equirectangular seam discontinuities."}, "model": "deepseek-v4-pro", "reviewedAt": "2026-09-10T00:52:28.558291Z", "state": "ai-reviewed"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "OmniSmall evaluates multi-object tracking of flying animals in omnidirectional video, focusing on seam-aware motion and association for small, fast-moving targets.", "displayEligible": true, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "public_reusable", "evidence": {"reasonCodes": ["benchmark term in abstract", "benchmark release stated in one sentence", "public artifact URL"], "snippet": "We propose a lightweight framework that re-engineers SORT-based tracking for this geometry, including (i) a Seam-Aware Motion Model that keeps the Kalman state continuous across the seam, (ii) a composite seam-aware association cost that pairs a wrapped Euclidean term with GIoU, and (iii) OmniSmall, a new benchmark of omnidirectional wildlife footage."}, "familyId": "bmf_33d267037bf8", "firstSeenAt": "2026-09-09", "heat": null, "id": "bm_re-engineering-sort-based-algorithms-for-l_55c64c2b", "indexedAt": "2026-09-09T06:18:29.794093Z", "industrySectors": ["Automotive"], "links": {"code": "https://github.com/Xin-Shu/OmniSORT.git", "data": null, "paper": "https://arxiv.org/abs/2609.07547", "pdf": "https://arxiv.org/pdf/2609.07547", "project": null, "report": "https://arxiv.org/abs/2609.07547"}, "metrics": [], "motivation": "Multi-object tracking (MOT) has advanced rapidly in urban surveillance and autonomous driving, yet many trackers rely on ReID- and transformer-based appearance encoders and are designed for standard FoV cameras.", "name": "OmniSmall", "oneLine": "OmniSmall evaluates multi-object tracking of flying animals in omnidirectional video, focusing on seam-aware motion and association for small, fast-moving targets.", "paperTitle": "Re-engineering SORT-based algorithms for low-cost small object tracking from omnidirectional footage", "primaryDomain": "Autonomous Driving", "publication": {"evidence": "Accepted by IEEE 28th International Workshop on Multimedia Signal Processing (MMSP)", "evidenceLevel": "author-claim", "evidenceUrl": "https://arxiv.org/abs/2609.07547", "source": "arxiv-comments", "status": "acceptance_claimed", "venue": "IEEE 28th International Workshop on Multimedia Signal Processing (MMSP)", "verifiedAt": "2026-09-09T06:18:29.794093Z"}, "publications": [], "publishers": [{"name": "Xin-Shu", "organizationType": "community", "role": "benchmark-publisher", "sourceUrl": "https://github.com/Xin-Shu/OmniSORT.git"}], "ranking": {"30d": {"confidence": "Low", "coverage": 0.55, "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": 0.10323645750017778, "value": 1}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.55, "rank": 102, "score": 28}, "method": "attention-ranking-v13", "rank": 102, "score": 28}, "90d": {"confidence": "Low", "coverage": 0.55, "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": 0.09361440452337855, "value": 1}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.55, "rank": 258, "score": 28}, "method": "attention-ranking-v13", "rank": 258, "score": 28}}, "readiness": "Runnable", "recognitionConfidence": 0.55, "relation": "introduces", "releasedAt": "2026-09-07", "source": {"authors": ["Xin Shu", "Meegan Gower", "Yvonne Buckley", "Anil Kokaram"], "categories": ["cs.CV"], "id": "2609.07547", "title": "Re-engineering SORT-based algorithms for low-cost small object tracking from omnidirectional footage", "type": "arxiv", "url": "https://arxiv.org/abs/2609.07547"}, "sourceUpdatedAt": "2026-09-07T14:29:21Z", "topics": ["cs.CV"], "venueAttempts": [{"decisionRaw": "Accepted by IEEE 28th International Workshop on Multimedia Signal Processing (MMSP)", "evidence": [{"level": "author-claim", "observedAt": "2026-09-09T06:18:29.794093Z", "rawValue": "Accepted by IEEE 28th International Workshop on Multimedia Signal Processing (MMSP)", "sourceType": "arxiv-comments", "sourceUrl": "https://arxiv.org/abs/2609.07547"}], "reviewStatus": "accepted", "venueName": "IEEE 28th International Workshop on Multimedia Signal Processing (MMSP)"}], "whyItMatters": "Addresses the need for tracking benchmarks that reflect low-cost omnidirectional deployments, where standard FoV assumptions fail due to projection seams and small target sizes."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
