{"id":"W3041972202","doi":"10.21203/rs.3.rs-19405/v1","title":"Automatic Facial Landmark Localization in Clinical Populations - Improving Model Performance with a Small Dataset","year":2020,"lang":"en","type":"preprint","venue":"","topic":"Facial Nerve Paralysis Treatment and Research","field":"Medicine","cited_by":9,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto; Toronto Rehabilitation Institute","funders":"Michael J. Fox Foundation for Parkinson's Research","keywords":"Landmark; Computer science; Artificial intelligence; Convolutional neural network; Population; Active appearance model; Pattern recognition (psychology); Computer vision; Image (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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00420398,0.001767346,0.0009969216,0.001029659,0.0005089318,0.001329012,0.001650698,0.001590993,0.002352041],"category_scores_gemma":[0.009320698,0.0004569821,0.001949197,0.0005943638,0.000431855,0.0009354157,0.001081821,0.0016594,0.001563081],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001273099,"about_ca_system_score_gemma":0.001116983,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01997506,"about_ca_topic_score_gemma":0.02245193,"domain_scores_codex":[0.9983879,0.0004139742,0.0001272226,0.0007555826,0.0001799913,0.0001353296],"domain_scores_gemma":[0.9970436,0.001392358,0.0001658594,0.0007744596,0.0005260861,0.00009776674],"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":[0.001590229,0.0007714473,0.07331144,0.0008169169,0.001384169,0.001019464,0.0002602438,0.3769674,0.02920397,0.0006253766,0.03961555,0.4744338],"study_design_scores_gemma":[0.0001097946,0.000258373,0.02128691,0.0001078575,0.0002262476,0.0003552432,0.0001335655,0.9595096,0.01210119,0.001038445,0.004798721,0.00007407481],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7887785,0.00470698,0.1726543,0.001700824,0.00128759,0.0007200672,0.01129245,0.01345892,0.005400523],"genre_scores_gemma":[0.8727348,0.0005978267,0.09741995,0.0006397567,0.0001444632,0.0004411904,0.02482669,0.000493542,0.002701763],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01997506,"threshold_uncertainty_score":0.03971761,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1969463998115574,"score_gpt":0.4069306761380942,"score_spread":0.2099842763265368,"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."}}