{"id":"W3031399851","doi":"10.1145/3383582","title":"Region-Level Visual Consistency Verification for Large-Scale Partial-Duplicate Image Search","year":2020,"lang":"en","type":"article","venue":"ACM Transactions on Multimedia Computing Communications and Applications","topic":"Advanced Image and Video Retrieval Techniques","field":"Computer Science","cited_by":37,"is_retracted":false,"has_abstract":true,"ca_institutions":"Lakehead University; University of Windsor","funders":"AUTO21 Network of Centres of Excellence; National Key Research and Development Program of China; National Natural Science Foundation of China","keywords":"Computer science; Artificial intelligence; Pattern recognition (psychology); Convolutional neural network; Consistency (knowledge bases); Pruning; Image retrieval; Image (mathematics); Feature (linguistics); Quantization (signal processing); Matching (statistics); Computer vision; Scale (ratio); Mathematics","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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","sts"],"consensus_categories":[],"category_scores_codex":[0.0003073089,0.0002432264,0.0002671081,0.0001318973,0.001540961,0.0001964232,0.002006244,0.00009536764,0.000004311248],"category_scores_gemma":[0.0001037901,0.0002617791,0.0001388485,0.0008393672,0.0002451951,0.0004265089,0.0001991764,0.0004270439,0.00004216806],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004673655,"about_ca_system_score_gemma":0.00009733197,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001595297,"about_ca_topic_score_gemma":0.000005839353,"domain_scores_codex":[0.9981018,0.0001198674,0.0005154505,0.0006953832,0.0001950082,0.000372447],"domain_scores_gemma":[0.9956204,0.001199676,0.0001738244,0.002352486,0.000385381,0.0002682379],"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":[0.00002715142,0.000686929,0.00007970195,0.00006220414,0.00005037469,4.932618e-7,0.002130399,0.0001168657,0.006187234,0.01202337,0.0002848076,0.9783505],"study_design_scores_gemma":[0.001091348,0.0002870097,0.0005225489,0.000051241,0.00006061075,0.0000145056,0.0005323018,0.9049222,0.01841361,0.002181702,0.07138669,0.0005361724],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0002202044,0.0002418879,0.9841291,0.01298176,0.000033271,0.001479634,0.00008893299,0.0006552104,0.0001700242],"genre_scores_gemma":[0.3760472,0.000694329,0.6218253,0.0006719709,0.0000570591,0.0005813753,0.00006822198,0.0000241676,0.00003045708],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.9778143,"threshold_uncertainty_score":0.9999834,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08380154559488553,"score_gpt":0.3572948206688599,"score_spread":0.2734932750739744,"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."}}