{"id":"W4320169013","doi":"10.1016/j.eswa.2023.119589","title":"A multi-task approach for contrastive learning of handwritten signature feature representations","year":2023,"lang":"en","type":"article","venue":"Expert Systems with Applications","topic":"Handwritten Text Recognition Techniques","field":"Computer Science","cited_by":23,"is_retracted":false,"has_abstract":false,"ca_institutions":"École de Technologie Supérieure; Université du Québec à Montréal","funders":"","keywords":"Computer science; Feature (linguistics); Signature (topology); Artificial intelligence; Task (project management); Feature learning; Feature vector; Pattern recognition (psychology); Class (philosophy); Natural language processing; 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":[],"consensus_categories":[],"category_scores_codex":[0.0002911324,0.0001799958,0.0002941999,0.0002517297,0.0003260181,0.0001150678,0.0005651374,0.0001482085,0.000001341511],"category_scores_gemma":[0.00006286416,0.0001499281,0.00008709138,0.001201521,0.00008122001,0.0002597263,0.00007971079,0.0002079649,0.00001437289],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000037067,"about_ca_system_score_gemma":0.00007650384,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00005531919,"about_ca_topic_score_gemma":0.000003055631,"domain_scores_codex":[0.9985111,0.00009156267,0.0002947741,0.000558403,0.0002664092,0.0002777776],"domain_scores_gemma":[0.9982976,0.0002983444,0.0002479928,0.0005555667,0.0005001412,0.0001003282],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0002368425,0.002467488,0.003406974,0.001859706,0.001529838,0.00002434328,0.0414772,0.0138434,0.3975109,0.2248932,0.2335327,0.07921743],"study_design_scores_gemma":[0.003340224,0.0004677033,0.001677883,0.0003499331,0.00006380101,0.0001061684,0.008402649,0.8526298,0.0273081,0.0006742399,0.1038617,0.001117833],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.000059842,0.0004949109,0.994161,0.0004433857,0.00003811392,0.002988997,0.00006007605,0.001004689,0.0007489537],"genre_scores_gemma":[0.4522105,0.00005081589,0.5200934,0.00007994458,0.000165071,0.02400238,0.0003884241,0.00004464211,0.002964812],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.8387864,"threshold_uncertainty_score":0.6113892,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02154042704118518,"score_gpt":0.2913953367677581,"score_spread":0.2698549097265729,"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."}}