{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2607.02271", "kind": "paper", "url": "https://arxiv.org/abs/2607.02271"}, {"id": "gh:advance-veintech-innovators/agvbench", "kind": "repo", "url": "https://github.com/Advance-VeinTech-Innovators/AGVBench"}], "authors": ["Haiyang Li", "Yuming Fu", "Qun Song", "Hongchao Liao", "Jing Chen", "Mounim A. EI-Yacoubi", "Yang Liu", "Siyuan Ma", "Xin Jin"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2607.02271", "source_url": "https://arxiv.org/abs/2607.02271"}, "categories": ["General AI", "Vision & 3D", "cs.CV"], "description": {"en": "AGVBench compares vein-recognition augmentation strategies across datasets and model families using accuracy, calibration, and adversarial-robustness considerations."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2607.02271", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2607.02271", "title": "AGVBench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["Chongqing Technology and Business University", "Chongqing University of Science and Technology", "Chongqing University of Technology"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2607.02271", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2607.02271", "source_url": "https://openalex.org/W7167269903"}, "key": "claire-radar:2607.02271", "modality": null, "name": "AGVBench", "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": "2026-09-25T04:40:22Z", "data_status": "primary-source-indexed", "display_eligible": "true", "origin_record_id": "bm_agvbench_b95fe35a", "origin_source": "arxiv", "recognition_confidence": "0.6", "record_sha256": "a0f54092d0c391317e7ea8426085b04e2ad956c20db4d08dc8bf5339d2cbe331", "relation": "introduces", "review_model": "codex-cli-default", "review_state": "ai-reviewed", "reviewed_at": "2026-09-24T19:01:37.943600Z", "source_url": "https://arxiv.org/abs/2607.02271"}, "publisher": null, "released": "2026-07-02", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2607.02271", "source_url": "https://arxiv.org/abs/2607.02271"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2607-02271", "source": "claire_radar", "source_benchmark_id": "2607.02271", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Vision & 3D", "attentionForecast": {"confidence": "Medium", "horizon": "7d", "reason": "The broad augmentation comparison and public implementation offer practical value, though vein recognition is a specialized application area.", "score": 57}, "capabilities": [], "confidence": "Medium", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-24T18:59:59.467212Z", "inputHash": "196032657e2e4d1174ec06262faf18c0af61f3c2f629f734224b3add6cb1aebc", "model": "codex-cli-default"}, "curation": {"admissionPolicyVersion": "2026-09-09.1", "canonicalNameEvidence": "AGVBench: A Reliability-Oriented Benchmark of Data Augmentation for Vein Recognition", "canonicalNameSource": "paper_title", "decisionReason": "The supplied evidence establishes a formal benchmark name, broad repeatable evaluation matrix, standardized protocols, scoring dimensions, and public codebase.", "evaluationEvidence": {"reuse": "Our codebase is available at https://github.com/Advance-VeinTech-Innovators/AGVBench.", "scoring": "clean accuracy and adversarial security", "task": "evaluates 30 representative augmentation strategies on five public palm- and finger-vein datasets with seven backbone architectures"}, "model": "codex-cli-default", "reviewedAt": "2026-09-24T19:01:37.943600Z", "state": "ai-reviewed"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "AGVBench compares vein-recognition augmentation strategies across datasets and model families using accuracy, calibration, and adversarial-robustness considerations.", "displayEligible": true, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "public_reusable", "evidence": {"reasonCodes": ["coined title prefix ending in Bench or Benchmark", "public artifact URL"], "snippet": "We present AGVBench, which evaluates 30 representative augmentation strategies on five public palm- and finger-vein datasets with seven backbone architectures, covering classic CNNs, vision transformers, and vein-specific recognition models."}, "familyId": "bmf_b5732b62654f", "firstSeenAt": "2026-08-19", "heat": null, "id": "bm_agvbench_b95fe35a", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": "https://github.com/Advance-VeinTech-Innovators/AGVBench", "data": null, "paper": "https://arxiv.org/abs/2607.02271", "pdf": "https://arxiv.org/pdf/2607.02271", "project": null, "report": "https://arxiv.org/abs/2607.02271"}, "metrics": [], "motivation": "Vein recognition is a secure biometric technology often constrained by limited annotated data and imaging variations.", "name": "AGVBench", "oneLine": "AGVBench compares vein-recognition augmentation strategies across datasets and model families using accuracy, calibration, and adversarial-robustness considerations.", "paperTitle": "AGVBench: A Reliability-Oriented Benchmark of Data Augmentation for Vein Recognition", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2607.02271", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-08-19T11:05:13.395391Z"}, "publications": [], "readiness": "Runnable", "recognitionConfidence": 0.6, "relation": "introduces", "releasedAt": "2026-07-02", "reviewLifecycle": {"state": "pending"}, "source": {"authors": ["Haiyang Li", "Yuming Fu", "Qun Song", "Hongchao Liao", "Jing Chen", "Mounim A.EI-Yacoubi", "Yang Liu", "Siyuan Ma", "Xin Jin"], "categories": ["cs.CV"], "id": "2607.02271", "title": "AGVBench: A Reliability-Oriented Benchmark of Data Augmentation for Vein Recognition", "type": "arxiv", "url": "https://arxiv.org/abs/2607.02271"}, "sourceUpdatedAt": "2026-07-26T09:45:17Z", "topics": ["cs.CV"], "venueAttempts": [], "whyItMatters": "It tests whether augmentation gains remain reliable and secure rather than optimizing clean recognition accuracy alone."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
