{"id":"W3196892597","doi":"10.48550/arxiv.2109.02137","title":"Efficient Action Recognition Using Confidence Distillation","year":2021,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Human Pose and Action Recognition","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"McMaster University","funders":"","keywords":"Computer science; Task (project management); Machine learning; Artificial intelligence; Representation (politics); Action recognition; Overhead (engineering); Aggregate (composite); Action (physics); Distillation; Sample (material); Pattern recognition (psychology); Artificial neural network; Engineering","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":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0001799156,0.0002277814,0.0002133035,0.0002612874,0.0002532371,0.0002947361,0.0003965178,0.0002535505,0.0001210134],"category_scores_gemma":[0.00003603087,0.0002946922,0.0001743972,0.000484092,0.00005086774,0.0003847636,0.0004916939,0.000416645,0.0001059776],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003344582,"about_ca_system_score_gemma":0.0001929119,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001533585,"about_ca_topic_score_gemma":0.00004094536,"domain_scores_codex":[0.998341,0.0001795285,0.0002096673,0.0009205669,0.0001211519,0.0002280806],"domain_scores_gemma":[0.9985671,0.00006270591,0.0003322196,0.0005660703,0.0003628141,0.0001090759],"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":[0.0000367085,0.0002527395,0.0003142633,0.0002129099,0.0001179328,0.0003370985,0.0005553152,0.9667631,0.001846174,0.01125462,0.00006372682,0.01824536],"study_design_scores_gemma":[0.0002306796,0.0000180075,0.0007360671,0.0002270624,0.00007001642,0.00001900208,0.0001654925,0.9857606,0.001473683,0.01089662,0.00004041804,0.0003623422],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4961386,0.00001444188,0.5024437,0.0000159844,0.0006724466,0.0001219214,0.000006528512,0.000134443,0.0004518817],"genre_scores_gemma":[0.9971471,0.00007544776,0.002344686,0.00005777727,0.0001323458,8.249197e-7,0.00009583912,0.00001215864,0.0001338599],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.5010085,"threshold_uncertainty_score":0.9999505,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2025058728526224,"score_gpt":0.2259309305121786,"score_spread":0.02342505765955616,"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."}}