{"id":"W2111141593","doi":"10.1109/iccv.2007.4408877","title":"Learning Structured Appearance Models from Captioned Images of Cluttered Scenes","year":2007,"lang":"en","type":"article","venue":"","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":20,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Computer science; Artificial intelligence; Feature (linguistics); ENCODE; Computer vision; Object (grammar); Graph; Pattern recognition (psychology)","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":[0.0002828978,0.0001240307,0.0001675269,0.0001022521,0.0001231643,0.00005356361,0.0006601784,0.00006086257,0.00002687639],"category_scores_gemma":[0.00004474822,0.0001148482,0.00005460819,0.0003050602,0.00005213659,0.00033375,0.0001580357,0.0002108477,0.00002978694],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002072231,"about_ca_system_score_gemma":0.00002226226,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002557508,"about_ca_topic_score_gemma":0.0000528558,"domain_scores_codex":[0.9988087,0.00005675187,0.0002899088,0.0003656775,0.0002505851,0.0002283342],"domain_scores_gemma":[0.9990394,0.0001135314,0.0001602535,0.0005046763,0.0001083148,0.00007387584],"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":[0.00004707286,0.0001301605,0.03880771,0.00003706941,0.00007280022,0.000006253855,0.00376404,0.2927342,0.3578676,0.09129555,0.0001132574,0.2151243],"study_design_scores_gemma":[0.0003783004,0.00002280361,0.2383031,0.00002038977,0.000005040743,0.000002537872,0.00005921661,0.7141126,0.0273841,0.01945952,0.00006539067,0.0001869484],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3293513,0.00006847508,0.6684361,0.0003081562,0.00004553147,0.00009005769,0.000001821151,0.0002073684,0.001491208],"genre_scores_gemma":[0.769229,0.000003897381,0.2305465,0.00005722906,0.00003719566,0.000004261898,0.000004847032,0.000008496112,0.0001086257],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.4398776,"threshold_uncertainty_score":0.4683375,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01213481241210248,"score_gpt":0.2609122937360986,"score_spread":0.2487774813239962,"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."}}