{"id":"W4391871827","doi":"10.48550/arxiv.2402.08532","title":"Captions Are Worth a Thousand Words: Enhancing Product Retrieval with Pretrained Image-to-Text Models","year":2024,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Mitacs","keywords":"Product (mathematics); Computer science; Image (mathematics); Natural language processing; Information retrieval; Artificial intelligence; Mathematics","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0004021652,0.0005212362,0.0004610248,0.0006107881,0.0003216013,0.0004547482,0.002032361,0.0002009193,0.00002546631],"category_scores_gemma":[0.0000940859,0.0005348248,0.0001860448,0.00228245,0.0001317566,0.0004028805,0.002817406,0.001627661,0.0002946622],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003669565,"about_ca_system_score_gemma":0.0005233525,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000449843,"about_ca_topic_score_gemma":0.0003146855,"domain_scores_codex":[0.9963756,0.0001682213,0.0002954504,0.002372406,0.0002328851,0.0005554011],"domain_scores_gemma":[0.9965626,0.0001266647,0.000298376,0.002320511,0.0003449808,0.0003468476],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00008036475,0.0001302155,0.0007319079,0.0002388848,0.0001773587,0.000261826,0.002119792,0.9387016,0.000485293,0.05628735,0.0002400567,0.000545353],"study_design_scores_gemma":[0.0003226938,0.00007015716,0.002423385,0.0005459877,0.0001545555,0.00001844571,0.0001266645,0.9548582,0.0002662586,0.04030299,0.0001205372,0.0007900542],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3297325,0.00006818911,0.665032,0.0009169194,0.000196667,0.0008837129,0.00003599059,0.000837034,0.002297025],"genre_scores_gemma":[0.9610539,0.00002491301,0.03602061,0.00008625404,0.000157974,0.0000125962,0.00002045221,0.00006490639,0.002558388],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.6313214,"threshold_uncertainty_score":0.9997103,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03925230774777872,"score_gpt":0.2080312265532537,"score_spread":0.168778918805475,"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."}}