{"id":"W4413155320","doi":"10.1109/cvpr52734.2025.01341","title":"BACON: Improving Clarity of Image Captions via Bag-of-Concept Graphs","year":2025,"lang":"en","type":"article","venue":"","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"CLARITY; Computer science; Artificial intelligence; Image (mathematics); Natural language processing; Information retrieval; Computer vision; Chemistry","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":[0.001377512,0.004266066,0.0009235884,0.003296222,0.0009363452,0.002476523,0.002975731,0.00284882,0.01389033],"category_scores_gemma":[0.008204624,0.001017983,0.002741213,0.001993597,0.001110681,0.004477616,0.002722164,0.00381617,0.007209098],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001805351,"about_ca_system_score_gemma":0.00143701,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01003288,"about_ca_topic_score_gemma":0.01496397,"domain_scores_codex":[0.9989429,0.0003187259,0.00005467702,0.0003845305,0.0002059267,0.00009323806],"domain_scores_gemma":[0.9973518,0.001255406,0.0001693754,0.0005199432,0.0006004312,0.0001030708],"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":[0.0007300134,0.0002267884,0.001632973,0.00251099,0.0002855798,0.0008992716,0.001217645,0.1280675,0.03360787,0.0213458,0.2688362,0.5406393],"study_design_scores_gemma":[0.0001066616,0.0001782117,0.0008272488,0.000325077,0.0001257152,0.0004674887,0.0002879143,0.8319322,0.02769678,0.02808036,0.1098358,0.0001364488],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01950897,0.003614696,0.8189214,0.001187508,0.001538499,0.0009582203,0.01325354,0.126224,0.01479314],"genre_scores_gemma":[0.1382313,0.001798726,0.7985816,0.001768533,0.0003588644,0.001158802,0.03646329,0.01055225,0.01108672],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01389033,"threshold_uncertainty_score":0.04646778,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.006728577190895381,"score_gpt":0.268057067877984,"score_spread":0.2613284906870886,"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."}}