{"id":"W4392864334","doi":"10.1145/3649896","title":"Realizing Efficient On-Device Language-based Image Retrieval","year":2024,"lang":"en","type":"article","venue":"ACM Transactions on Multimedia Computing Communications and Applications","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto; Centre for Social Innovation","funders":"","keywords":"Computer science; Information retrieval; Image retrieval; Latency (audio); Ranking (information retrieval); Modal; Language model; Deep learning; Context (archaeology); Artificial intelligence; Image (mathematics)","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":[0.0005182289,0.0009224473,0.001145166,0.0006769794,0.0004294962,0.001244331,0.002182275,0.001238972,0.00678025],"category_scores_gemma":[0.001880572,0.0004289025,0.0009239783,0.0007249636,0.0005761648,0.003263091,0.00216025,0.0009507205,0.005752987],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000611377,"about_ca_system_score_gemma":0.0007603217,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003790833,"about_ca_topic_score_gemma":0.007415783,"domain_scores_codex":[0.999424,0.00006715594,0.00003232374,0.0001405576,0.0002165137,0.0001196026],"domain_scores_gemma":[0.9995313,0.0001225419,0.00003428746,0.0001555412,0.0001223036,0.00003408987],"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":[0.001020634,0.0006603946,0.001362461,0.0008315203,0.000169474,0.0008338048,0.0003895752,0.08341663,0.3061387,0.02300074,0.0396106,0.5425655],"study_design_scores_gemma":[0.00007129466,0.0001872327,0.0004209297,0.000022918,0.00004913834,0.000402245,0.0001150943,0.8987833,0.07817357,0.01192346,0.009787461,0.00006333096],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.07013846,0.002193875,0.8932771,0.0005339652,0.0002115662,0.0002715264,0.0007735129,0.01670874,0.01589126],"genre_scores_gemma":[0.6289086,0.0009317953,0.3499444,0.0008870829,0.0001261115,0.0002605522,0.001937902,0.0007187404,0.01628477],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.00678025,"threshold_uncertainty_score":0.02268213,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02008231110073879,"score_gpt":0.3266956285550984,"score_spread":0.3066133174543596,"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."}}