{"id":"W3034733309","doi":"10.18653/v1/2020.acl-main.664","title":"Improving Image Captioning with Better Use of Caption","year":2020,"lang":"en","type":"article","venue":"","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":85,"is_retracted":false,"has_abstract":true,"ca_institutions":"Queen's University","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Closed captioning; Computer science; Leverage (statistics); Artificial intelligence; Natural language processing; Semantics (computer science); Visualization; Feature learning; Representation (politics); Word (group theory); Inductive bias; Image (mathematics); Machine learning; Task (project management); Multi-task learning","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00188524,0.002661191,0.001286784,0.002232692,0.0008249556,0.002913401,0.002826238,0.002700149,0.007712781],"category_scores_gemma":[0.00989101,0.0006015322,0.001663821,0.002069376,0.001161255,0.00497949,0.002813195,0.003596728,0.006107907],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001218828,"about_ca_system_score_gemma":0.0008954412,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002669281,"about_ca_topic_score_gemma":0.003491987,"domain_scores_codex":[0.9985659,0.0004609476,0.00006705638,0.000509947,0.0002896257,0.0001066237],"domain_scores_gemma":[0.9966292,0.001312143,0.0002653268,0.000988109,0.000664509,0.0001408096],"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":[0.0004307994,0.0003542585,0.001200207,0.0009875093,0.0001914294,0.0004358585,0.0004623768,0.09661594,0.04339142,0.01212291,0.1002077,0.7435996],"study_design_scores_gemma":[0.00007124173,0.0001726436,0.0005741382,0.0001104485,0.0001126344,0.0004724415,0.0001534718,0.8922347,0.04686318,0.02222504,0.03694197,0.00006809171],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01985859,0.003570586,0.9272242,0.00144389,0.001056104,0.0006021062,0.002563799,0.03160476,0.01207593],"genre_scores_gemma":[0.2081635,0.002005955,0.761292,0.001432993,0.0006837877,0.0004734293,0.01260062,0.003289548,0.01005808],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007712781,"threshold_uncertainty_score":0.02580184,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02151848015296971,"score_gpt":0.2306801638018164,"score_spread":0.2091616836488467,"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."}}