{"id":"W3193719728","doi":"10.1145/3462244.3479955","title":"Self-supervised Contrastive Learning of Multi-view Facial Expressions","year":2021,"lang":"en","type":"preprint","venue":"","topic":"Emotion and Mood Recognition","field":"Psychology","cited_by":32,"is_retracted":false,"has_abstract":true,"ca_institutions":"Queen's University","funders":"","keywords":"Computer science; Robustness (evolution); Artificial intelligence; Facial expression; Encoder; Invariant (physics); Pattern recognition (psychology); Exploit; Training set; Labeled data; Face (sociological concept); Component (thermodynamics); Speech recognition; Computer vision; Mathematics","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":"codex-gemma-dda1882f352a","candidate_categories":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0001794336,0.0002288689,0.0004742391,0.0001096824,0.00007215675,0.00002622336,0.0001489103,0.0005113823,0.01203194],"category_scores_gemma":[0.000108781,0.000211647,0.000266179,0.00007863662,0.00004485881,0.00003407546,0.000254215,0.0008599786,0.0001843981],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002440533,"about_ca_system_score_gemma":0.000101744,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001527764,"about_ca_topic_score_gemma":0.00004215056,"domain_scores_codex":[0.9981942,0.0005171766,0.0004329254,0.0004878855,0.0001524362,0.0002153314],"domain_scores_gemma":[0.9990101,0.0001141474,0.0002270878,0.0002685699,0.0002803584,0.00009967357],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"qualitative","study_design_scores_codex":[0.0004362826,0.01179412,0.01267124,0.003833284,0.005728494,0.0003037333,0.1498123,0.0005877744,0.02793163,0.00693886,0.0068653,0.773097],"study_design_scores_gemma":[0.06826907,0.002376348,0.280997,0.01913009,0.005763485,0.000321036,0.3069489,0.06290432,0.05151134,0.001814637,0.1875602,0.01240359],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.4101523,0.00372337,0.1827409,0.0004602283,0.008667327,0.002458761,0.0002049505,0.001069459,0.3905227],"genre_scores_gemma":[0.9752431,0.000425221,0.01458638,0.0001996544,0.0001751974,0.0001489312,0.0005101488,0.00004000988,0.008671349],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.7606934,"threshold_uncertainty_score":0.9888712,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06554403347052203,"score_gpt":0.3495080473286774,"score_spread":0.2839640138581553,"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."}}