{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2606.27128", "kind": "paper", "url": "https://arxiv.org/abs/2606.27128"}], "authors": ["Mobin Habibpour", "John Spodnik", "Niloufar Alipour Talemi", "Fatemeh Afghah"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2606.27128", "source_url": "https://arxiv.org/abs/2606.27128"}, "categories": ["General AI", "Reasoning", "Safety", "Safety & Trustworthiness"], "description": {"en": "FlameVQA is a multiple-choice visual question answering benchmark for UAV-based wildfire monitoring, built on FLAME 3 with paired RGB and radiometric thermal images. It includes 34 questions per image across six capability groups, covering detection, localization, coverage estimation, cross-modal reasoning, and flight planning. Labels are generated via MLLM assistance, deterministic thermal rules, and human auditing. The dataset and code are open-source."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2606.27128", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2606.27128", "title": "FlameVQA"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2606.27128", "modality": null, "name": "FlameVQA", "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_flamevqa_fd5c1a64", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "2f71367d863778dbc82dd3c1849e8c7667f8d9399c322fa0c53bf64b09cc981e", "relation": "introduces", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2606.27128"}, "publisher": null, "released": "2026-06-25", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2606.27128", "source_url": "https://arxiv.org/abs/2606.27128"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2606-27128", "source": "claire_radar", "source_benchmark_id": "2606.27128", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Safety & Trustworthiness", "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/2606.27128", "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": ["Reasoning"], "confidence": "High", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-08-20T15:22:10.044731Z", "inputHash": "a632b6de7e18eb2936658e9aa29a3bd19fe205b989d42a1ee97d624c48417fea", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "FlameVQA is a multiple-choice visual question answering benchmark for UAV-based wildfire monitoring, built on FLAME 3 with paired RGB and radiometric thermal images. It includes 34 questions per image across six capability groups, covering detection, localization, coverage estimation, cross-modal reasoning, and flight planning. Labels are generated via MLLM assistance, deterministic thermal rules, and human auditing. The dataset and code are open-source.", "domainCuration": {"method": "rules-v1", "state": "auto"}, "evidence": {"reasonCodes": ["exact coined title identity tied to benchmark evidence", "benchmark release stated in one sentence", "evaluation protocol evidence"], "snippet": "We introduce FlameVQA, a multiple-choice visual question answering benchmark for UAV-based wildfire intelligence built on FLAME 3, leveraging paired RGB imagery and radiometric thermal TIFFs for temperature-grounded, safety-critical reasoning."}, "familyId": "bmf_5e31ad44d17e", "firstSeenAt": "2026-08-19", "githubIndex": {"candidateCount": 1, "indexedAt": "2026-09-17T04:40:34.111665+00:00", "repositories": [], "status": "candidates"}, "heat": null, "id": "bm_flamevqa_fd5c1a64", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": null, "data": null, "hfPaper": "https://huggingface.co/papers/2606.27128", "paper": "https://arxiv.org/abs/2606.27128", "pdf": "https://arxiv.org/pdf/2606.27128", "project": null, "report": "https://arxiv.org/abs/2606.27128"}, "metrics": [], "motivation": "Wildfire monitoring from UAVs requires reliable reasoning over complex aerial scenes, where smoke, scale variation, and occlusions often limit RGB-only interpretation.", "name": "FlameVQA", "oneLine": "FlameVQA is a multiple-choice visual question answering benchmark for UAV-based wildfire monitoring, built on FLAME 3 with paired RGB and radiometric thermal images. It includes 34 questions per image across six capability groups, covering detection, localization, coverage estimation, cross-modal reasoning, and flight planning. Labels are generated via MLLM assistance, deterministic thermal rules, and human auditing. The dataset and code are open-source.", "paperTitle": "FlameVQA: A Physically-Grounded UAV Wildfire VQA Benchmark with Radiometric Thermal Supervision", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2606.27128", "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": "2026-06-25", "source": {"authors": ["Mobin Habibpour", "John Spodnik", "Niloufar Alipour Talemi", "Fatemeh Afghah"], "categories": ["cs.CV", "cs.RO"], "id": "2606.27128", "title": "FlameVQA: A Physically-Grounded UAV Wildfire VQA Benchmark with Radiometric Thermal Supervision", "type": "arxiv", "url": "https://arxiv.org/abs/2606.27128"}, "sourceUpdatedAt": "2026-06-25T15:06:59Z", "topics": ["Safety", "Reasoning"], "venueAttempts": [], "whyItMatters": "FlameVQA addresses the lack of benchmarks for evaluating vision-language models in safety-critical wildfire scenarios where RGB-only interpretation is insufficient. It provides a standardized evaluation for capabilities like smoke detection and coverage estimation, which are critical for practical deployment of MLLMs in disaster monitoring."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
