{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001906699,0.002518443,0.001225724,0.004348123,0.000892844,0.002038295,0.003758176,0.002603214,0.01972322],"category_scores_gemma":[0.006390942,0.0007797802,0.00244759,0.001816134,0.0009334445,0.00587007,0.005931118,0.003495964,0.008593424],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001389273,"about_ca_system_score_gemma":0.0016798,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0108472,"about_ca_topic_score_gemma":0.019304,"domain_scores_codex":[0.9986873,0.0003189505,0.00008118897,0.0005527477,0.0002352052,0.0001245915],"domain_scores_gemma":[0.9984035,0.0006302122,0.0000708567,0.0005234199,0.0002651374,0.0001068606],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0003750665,0.0005455362,0.001954983,0.001032236,0.000327181,0.0002374527,0.0007214788,0.01787088,0.01792385,0.01200623,0.05640797,0.8905971],"study_design_scores_gemma":[0.0001520491,0.0003263395,0.003239434,0.0002949658,0.0002987863,0.000429404,0.0008415524,0.8095008,0.0285896,0.07887284,0.07733202,0.0001222094],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01224932,0.001573388,0.913879,0.0005336048,0.0001788026,0.0006976005,0.005000492,0.06026256,0.005625213],"genre_scores_gemma":[0.1343512,0.0006785998,0.8344306,0.0006689911,0.00009332745,0.0007664847,0.02143489,0.0008598075,0.006716046],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01972322,"threshold_uncertainty_score":0.06598067,"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."}}