{"id":"W3034455598","doi":"10.24963/ijcai.2020/285","title":"Compressed Self-Attention for Deep Metric Learning with Low-Rank Approximation","year":2020,"lang":"en","type":"article","venue":"","topic":"Video Surveillance and Tracking Methods","field":"Computer Science","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"Artificial Intelligence in Medicine (Canada)","funders":"Wuhan University; Fundamental Research Funds for the Central Universities; Natural Science Foundation of Hubei Province; National Natural Science Foundation of China","keywords":"Metric (unit); Computer science; Pooling; Landmark; Artificial intelligence; Feature (linguistics); Pairwise comparison; Rank (graph theory); Deep learning; Pattern recognition (psychology); Computation; Similarity (geometry); Task (project management); Machine learning; Algorithm; Mathematics; 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":[0.001020195,0.001288245,0.001239587,0.0008179745,0.0003620716,0.0007553254,0.002023201,0.001148929,0.003766855],"category_scores_gemma":[0.004243676,0.0004170734,0.0006599989,0.001125227,0.0007189798,0.002142707,0.001700241,0.001702305,0.001322177],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001094993,"about_ca_system_score_gemma":0.001198225,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.007506793,"about_ca_topic_score_gemma":0.009277217,"domain_scores_codex":[0.9991886,0.0001899792,0.00004082555,0.0001959116,0.0002752002,0.0001094527],"domain_scores_gemma":[0.9990396,0.0003290735,0.00009774614,0.0002443098,0.0002355838,0.00005359302],"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":[0.0002276878,0.000210757,0.00111997,0.000196646,0.0001130098,0.0001288107,0.000133993,0.2670848,0.01478818,0.02219428,0.01427756,0.6795244],"study_design_scores_gemma":[0.000007986873,0.00004457465,0.0001359265,0.000004128251,0.000006861511,0.00002793332,0.000006568851,0.9902807,0.002530663,0.006022611,0.0009254096,0.000006606408],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01228494,0.000559663,0.9828011,0.0002299504,0.00006197366,0.00003588036,0.0001229841,0.002844712,0.001058755],"genre_scores_gemma":[0.5860898,0.0006220057,0.4026995,0.000636713,0.0002721954,0.0002108982,0.001319897,0.0004433335,0.007705631],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007506793,"threshold_uncertainty_score":0.0149262,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02014887226945844,"score_gpt":0.2600178486111484,"score_spread":0.2398689763416899,"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."}}