{"id":"W3132507547","doi":"10.1016/j.patcog.2022.109209","title":"Visual question answering from another perspective: CLEVR mental rotation tests","year":2022,"lang":"en","type":"article","venue":"Pattern Recognition","topic":"Advanced Image and Video Retrieval Techniques","field":"Computer Science","cited_by":6,"is_retracted":false,"has_abstract":false,"ca_institutions":"Canadian Institute for Advanced Research; McGill University; Minnow Environmental (Canada); Polytechnique Montréal; Mila - Quebec Artificial Intelligence Institute","funders":"Institut de Valorisation des Données; Natural Sciences and Engineering Research Council of Canada; Mitacs; Artificial Intelligence Research Center; Canadian Institute for Advanced Research","keywords":"Mental rotation; Perspective (graphical); Artificial intelligence; Computer science; Mental image; Rotation (mathematics); Perception; Object (grammar); Question answering; Computer vision; Visual perception; Cognitive psychology; Psychology; Cognition","routes":{"ca_aff":true,"ca_fund":true,"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":[0.0001650347,0.0001126267,0.00009086399,0.0001077558,0.0002320701,0.00008834683,0.0002004616,0.00002640478,0.0002477984],"category_scores_gemma":[0.0000291792,0.0001291832,0.00004586826,0.0002329677,0.0000153686,0.0008234796,0.0001650209,0.0001649976,0.00006099424],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002645639,"about_ca_system_score_gemma":0.0000160717,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003034142,"about_ca_topic_score_gemma":0.00001508414,"domain_scores_codex":[0.9989034,0.0001395217,0.0001555953,0.0003671841,0.0002767513,0.0001575825],"domain_scores_gemma":[0.999585,0.00004813227,0.0001052652,0.0001489086,0.00008212988,0.00003057769],"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":[0.00001866769,0.0001227677,0.002242923,0.0000038532,0.00001114377,0.00001386252,0.001393271,0.000006423711,0.02733297,0.00004550145,0.0001734685,0.9686351],"study_design_scores_gemma":[0.002639028,0.002142713,0.09439066,0.0002300408,0.00005444076,0.000131321,0.003913894,0.0814872,0.6943674,0.114201,0.004582583,0.001859747],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1762675,0.00008039165,0.8221266,0.0003381539,0.0002412555,0.0002267153,0.00004196829,0.0003649981,0.0003123931],"genre_scores_gemma":[0.9845046,0.00002690196,0.01439634,0.000678343,0.0001100048,0.0001079413,0.0001296677,0.00001693472,0.00002921128],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9667754,"threshold_uncertainty_score":0.5267938,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02354854623864443,"score_gpt":0.3167079208667227,"score_spread":0.2931593746280782,"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."}}