{"id":"W4221155360","doi":"10.1109/cvpr52688.2022.00503","title":"MuKEA: Multimodal Knowledge Extraction and Accumulation for Knowledge-based Visual Question Answering","year":2022,"lang":"en","type":"article","venue":"2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":116,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal","funders":"National Natural Science Foundation of China","keywords":"Computer science; Embedding; Knowledge extraction; Construct (python library); Domain knowledge; Question answering; Pipeline (software); Relation (database); Commonsense knowledge; Knowledge base; Artificial intelligence; Domain (mathematical analysis); Bridge (graph theory); Natural language processing; Data mining","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":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0007505734,0.0003494691,0.0003194645,0.0004510579,0.0009597972,0.0004408958,0.0003407559,0.0001225089,0.000136582],"category_scores_gemma":[0.0000359179,0.0003767992,0.00009791808,0.0003551563,0.00005932809,0.0005189896,0.0002685702,0.000544189,0.00006086493],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001212741,"about_ca_system_score_gemma":0.0001017516,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001088998,"about_ca_topic_score_gemma":0.00004015421,"domain_scores_codex":[0.9973435,0.0005316333,0.0004614732,0.001012059,0.0003157188,0.000335585],"domain_scores_gemma":[0.9981999,0.000692323,0.0002794839,0.0003452573,0.0002907616,0.0001922204],"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":[0.00008627146,0.0005197531,0.0008585485,0.00009068382,0.0000173579,0.000003479862,0.0007407886,0.001861133,0.003872856,0.0005009706,0.0002989216,0.9911492],"study_design_scores_gemma":[0.00166736,0.0009836555,0.01862545,0.0001184405,0.00001900683,0.00003182911,0.00004761305,0.9752714,0.0007562657,0.0007486386,0.00128644,0.0004439069],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3285006,0.00003887999,0.6688764,0.0007310643,0.0007533727,0.0006632897,0.00004130839,0.000255765,0.0001393436],"genre_scores_gemma":[0.9798418,0.00002087996,0.0185441,0.0004319355,0.0002611229,0.0005071588,0.0002903282,0.00003430702,0.00006838775],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9907053,"threshold_uncertainty_score":0.9998684,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06767699411909972,"score_gpt":0.3792622694143057,"score_spread":0.311585275295206,"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."}}