{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2609.10394", "kind": "paper", "url": "https://arxiv.org/abs/2609.10394"}], "authors": ["Rishabh Jain", "Aristeidis Papadopoulos", "Zhaofeng Lin", "Naomi Harte"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2609.10394", "source_url": "https://arxiv.org/abs/2609.10394"}, "categories": ["General AI", "Multimodal", "eess.AS"], "description": {"en": "Audio-visual speech recognition on natural dyadic videoconference conversations."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2609.10394", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2609.10394", "title": "Candor-LR"}], "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.2609.10394", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2609.10394", "source_url": "https://openalex.org/W7212114199"}, "key": "claire-radar:2609.10394", "modality": null, "name": "Candor-LR", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"admission_policy_version": "2026-09-09.1", "confidence": "High", "crawl_bundle": "claire_radar_2026-10-14", "crawled_at": "2026-10-14T09:44:51Z", "data_status": "primary-source-indexed", "display_eligible": "false", "origin_record_id": "bm_candor-lr_c4d723cf", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "986e78c2a1e71e446b978181ff8fb45e8a2e70e5549bb86fbaf966a79e93240d", "relation": "introduces", "review_model": "deepseek-v4-pro", "review_state": "ai-name-audit-deferred", "reviewed_at": "2026-09-10T09:05:59.100994Z", "source_url": "https://arxiv.org/abs/2609.10394"}, "publisher": null, "released": "2026-09-09", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2609.10394", "source_url": "https://arxiv.org/abs/2609.10394"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2609-10394", "source": "claire_radar", "source_benchmark_id": "2609.10394", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Multimodal", "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/2609.10394", "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"}, "attentionForecast": {"confidence": "Low", "horizon": "7d", "reason": "Conversational speech recognition is an active area, but missing direct artifacts limits visibility.", "score": 25}, "capabilities": [], "confidence": "High", "construction": "Transform Existing", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-10T09:05:59.100994Z", "inputHash": "856f819d38cf53680aa5d7f2697b8119cc54748d84834a533d5cbbc8262e4e22", "model": "deepseek-v4-pro"}, "curation": {"admissionPolicyVersion": "2026-09-09.1", "canonicalNameEvidence": "we introduce Candor-LR, a conversational benchmark derived from the CANDOR corpus of 1,656 natural dyadic videoconferences.", "canonicalNameSource": "abstract", "decisionReason": "Although a pipeline is mentioned as open-source, no code or dataset link is provided in the evidence.", "evaluationEvidence": {"reuse": "We open-source our pipeline to ensure reproducibility, establishing Candor-LR as a challenging benchmark for conversational AVSR.", "scoring": "Evaluating pretrained AVSR models on Candor-LR reveals that audio-only accuracy drops sharply compared to LRS3, but visual cues compensate effectively.", "task": "Candor-LR is a conversational benchmark for audio-visual speech recognition, derived from natural dyadic videoconferences with training, validation, and test splits."}, "model": "deepseek-v4-pro", "reviewedAt": "2026-09-10T09:05:59.100994Z", "state": "ai-name-audit-deferred"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "Audio-visual speech recognition on natural dyadic videoconference conversations.", "displayEligible": false, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "public_reusable", "evidence": {"reasonCodes": ["exact coined title identity tied to benchmark evidence", "benchmark release stated in one sentence"], "snippet": "To shift the field toward realistic dialogue, we introduce Candor-LR, a conversational benchmark derived from the CANDOR corpus of 1,656 natural dyadic videoconferences."}, "familyId": "bmf_fe67b0d002ee", "firstSeenAt": "2026-09-10", "heat": null, "id": "bm_candor-lr_c4d723cf", "indexedAt": "2026-09-10T06:16:58.931316Z", "industrySectors": [], "links": {"code": null, "data": null, "paper": "https://arxiv.org/abs/2609.10394", "pdf": "https://arxiv.org/pdf/2609.10394", "project": null, "report": "https://arxiv.org/abs/2609.10394"}, "metrics": [], "motivation": "Current audio-visual speech recognition (AVSR) benchmarks, like LRS3, rely heavily on clean, scripted and rehearsed speech.", "name": "Candor-LR", "oneLine": "Audio-visual speech recognition on natural dyadic videoconference conversations.", "paperTitle": "Candor-LR: A Dyadic Conversational Dataset for Audio-Visual Speech Recognition", "primaryDomain": "General AI", "publication": {"evidence": "Accepted to IEEE SLT 2026", "evidenceLevel": "author-claim", "evidenceUrl": "https://arxiv.org/abs/2609.10394", "source": "arxiv-comments", "status": "acceptance_claimed", "venue": "IEEE SLT 2026", "verifiedAt": "2026-09-10T06:16:58.931316Z"}, "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": 30, "historyDate": "2026-09-05", "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": 1.0, "relation": "introduces", "releasedAt": "2026-09-09", "source": {"authors": ["Rishabh Jain and Aristeidis Papadopoulos and Zhaofeng Lin and Naomi Harte"], "categories": ["eess.AS", "cs.CV", "cs.MM"], "id": "2609.10394", "title": "Candor-LR: A Dyadic Conversational Dataset for Audio-Visual Speech Recognition", "type": "arxiv", "url": "https://arxiv.org/abs/2609.10394"}, "sourceUpdatedAt": "2026-09-09T16:14:29Z", "topics": ["eess.AS"], "venueAttempts": [{"decisionRaw": "Accepted to IEEE SLT 2026", "evidence": [{"level": "author-claim", "observedAt": "2026-09-10T06:16:58.931316Z", "rawValue": "Accepted to IEEE SLT 2026", "sourceType": "arxiv-comments", "sourceUrl": "https://arxiv.org/abs/2609.10394"}], "reviewStatus": "accepted", "venueName": "IEEE SLT 2026"}], "whyItMatters": "Clean scripted benchmarks like LRS3 fail to capture overlapping speech and spontaneous turn-taking in realistic dialogue."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
