{"id":"W7108068765","doi":"10.5281/zenodo.17770659","title":"Robust Multimodal Engagement Recognition Using Emotion–Gaze–Pose Fusion with Lightweight LLM Temporal Reasoning","year":2025,"lang":"en","type":"preprint","venue":"Zenodo (CERN European Organization for Nuclear Research)","topic":"Emotion and Mood Recognition","field":"Psychology","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Artificial Intelligence in Medicine (Canada)","funders":"","keywords":"Upload; Gaze; Action (physics); Baseline (sea); Code (set theory); Action recognition; Fusion; Emotion recognition","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0004316171,0.000919237,0.0005839078,0.0009833365,0.0002287667,0.0009553991,0.0007168463,0.0004900246,0.007225686],"category_scores_gemma":[0.001570315,0.0001879112,0.0006436884,0.0005199314,0.0001805115,0.0008807342,0.001680383,0.0006090727,0.004143252],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000351172,"about_ca_system_score_gemma":0.000381415,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002189725,"about_ca_topic_score_gemma":0.00449693,"domain_scores_codex":[0.999649,0.00004636499,0.00001596051,0.0001324173,0.0001018246,0.00005444926],"domain_scores_gemma":[0.9997526,0.00005429586,0.00003075714,0.00005756259,0.0000754103,0.00002940391],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0009393203,0.0002122774,0.005407988,0.0003466633,0.0001577919,0.000229583,0.0002725215,0.009529844,0.1615375,0.002148246,0.0195453,0.7996728],"study_design_scores_gemma":[0.00009100929,0.0005247544,0.03156257,0.0001506335,0.0001748716,0.0007288138,0.0003875152,0.778029,0.1347879,0.01893121,0.03448504,0.0001466989],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.03849389,0.0007806734,0.9265553,0.0002473105,0.0002070041,0.0001883953,0.004665132,0.02408397,0.004778313],"genre_scores_gemma":[0.4857858,0.0007384226,0.4879764,0.0004374876,0.0002677507,0.00080195,0.01326676,0.001308028,0.009417407],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007225686,"threshold_uncertainty_score":0.02417237,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1110034961556459,"score_gpt":0.3036817796873797,"score_spread":0.1926782835317338,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}