{"id":"W4402728032","doi":"10.1109/cvpr52733.2024.00386","title":"Emergent Open-Vocabulary Semantic Segmentation from Off-the-Shelf Vision-Language Models","year":2024,"lang":"en","type":"article","venue":"","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":12,"is_retracted":false,"has_abstract":true,"ca_institutions":"Vector Institute; University of British Columbia","funders":"","keywords":"Computer science; Vocabulary; Segmentation; Natural language processing; Artificial intelligence; Image segmentation; Linguistics","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":[0.0003066964,0.0001415925,0.0001127973,0.00007432597,0.0002099558,0.0008462602,0.001605202,0.00004013772,0.0002392074],"category_scores_gemma":[0.00001367785,0.0001001397,0.00006006783,0.0004468101,0.00002377468,0.0008430154,0.0007854649,0.0001981048,0.0007655235],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004421937,"about_ca_system_score_gemma":0.00005765092,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004147834,"about_ca_topic_score_gemma":0.00008436532,"domain_scores_codex":[0.998608,0.0001016507,0.0002484107,0.0005493813,0.0003015717,0.0001909911],"domain_scores_gemma":[0.9988083,0.0001889175,0.00004247463,0.0008594596,0.0000237273,0.00007708641],"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":[0.000007443019,0.0002353023,0.0002589752,0.00003182823,0.0001554611,0.00004778921,0.01181377,0.03559273,0.03216125,0.2031025,0.0171597,0.6994333],"study_design_scores_gemma":[0.000118407,0.00002218272,0.001271153,0.0000235819,0.00001304642,0.000004396067,0.0001011393,0.9821598,0.001378855,0.01326717,0.001500506,0.0001397619],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.05557486,0.001009891,0.9276848,0.01053616,0.0003400879,0.0005050644,0.000005692308,0.0004772102,0.003866246],"genre_scores_gemma":[0.9159917,0.00003815377,0.081742,0.0007819639,0.00007731047,0.000121167,0.00004425726,0.00001832189,0.001185182],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9465671,"threshold_uncertainty_score":0.983951,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01812614014927617,"score_gpt":0.3248814039916032,"score_spread":0.306755263842327,"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."}}