{"id":"W4391173394","doi":"10.1117/1.jei.33.1.013028","title":"TEG: image theme recognition using text-embedding-guided few-shot adaptation","year":2024,"lang":"en","type":"article","venue":"Journal of Electronic Imaging","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Memorial University of Newfoundland","funders":"","keywords":"Computer science; Artificial intelligence; Computer vision; Theme (computing); Image processing; Embedding; Adaptation (eye); Pattern recognition (psychology); Image (mathematics); Optics; Physics","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":[0.001038392,0.0001612683,0.0001812328,0.0004171758,0.00017094,0.000546079,0.0005207742,0.00003310293,0.00003828116],"category_scores_gemma":[0.0001293659,0.0001473243,0.0001512438,0.0006128165,0.00003178363,0.001511408,0.0000732569,0.0007477925,0.00005340539],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004892609,"about_ca_system_score_gemma":0.0005477462,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001114651,"about_ca_topic_score_gemma":0.000003588149,"domain_scores_codex":[0.9983097,0.0001163043,0.0005095525,0.0002724086,0.0003657383,0.0004263375],"domain_scores_gemma":[0.9989027,0.0001801483,0.0003233216,0.0002394619,0.0002755763,0.00007872892],"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":[0.00003237358,0.000185138,0.0004350377,0.0001326188,0.0002425556,0.0001833551,0.00422339,0.03437508,0.3456748,0.02221079,0.002107447,0.5901974],"study_design_scores_gemma":[0.0002362066,0.00004640114,0.0001875911,0.00016137,0.00004234599,0.001191414,0.00006542436,0.9785334,0.002519347,0.01527641,0.001573652,0.0001664779],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.09085244,0.001337291,0.9027542,0.003661853,0.0002902855,0.0001169617,8.849469e-7,0.0001459483,0.0008401683],"genre_scores_gemma":[0.8844052,0.00004030158,0.115053,0.0001475074,0.0002897446,0.000003989222,0.00000240961,0.00002565462,0.000032208],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9441583,"threshold_uncertainty_score":0.6007712,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03676406548906194,"score_gpt":0.3436573871034038,"score_spread":0.3068933216143419,"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."}}