{"id":"W3208470653","doi":"10.1109/iccvw54120.2021.00091","title":"Knowledge Distillation for Low-Power Object Detection: A Simple Technique and Its Extensions for Training Compact Models Using Unlabeled Data","year":2021,"lang":"en","type":"article","venue":"","topic":"Advanced Neural Network Applications","field":"Computer Science","cited_by":12,"is_retracted":false,"has_abstract":true,"ca_institutions":"Huawei Technologies (Canada)","funders":"","keywords":"Distillation; Computer science; Simple (philosophy); Artificial intelligence; Object (grammar); Training set; Training (meteorology); Machine learning; Object detection; Power (physics); Pattern recognition (psychology); Chromatography; Chemistry","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":[],"consensus_categories":[],"category_scores_codex":[0.000195795,0.0001413573,0.0001930131,0.00005840475,0.0004500405,0.00009238307,0.0003413229,0.00006077005,0.000002897486],"category_scores_gemma":[0.0001351072,0.0001382512,0.00004277538,0.0004910897,0.00002177437,0.0007988315,0.0002781726,0.00007617771,5.854956e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004338344,"about_ca_system_score_gemma":0.000113778,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000001997711,"about_ca_topic_score_gemma":0.00004816328,"domain_scores_codex":[0.998755,0.00003253896,0.0002395435,0.0006299958,0.00007806663,0.0002649029],"domain_scores_gemma":[0.9982613,0.0005556604,0.00008608606,0.0007132821,0.0002928902,0.00009077119],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00006107664,0.0002590818,0.0000104009,0.000155923,0.00008529955,0.000004908561,0.001208386,0.08301364,0.6394676,0.1826201,0.001119909,0.09199371],"study_design_scores_gemma":[0.0002913289,0.00003349167,0.00001891109,0.00002308754,0.00001393197,0.00004495788,0.00003792558,0.9453591,0.02580304,0.02667174,0.001534119,0.0001683462],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.005151974,0.0002190912,0.9927674,0.0002134465,0.0000775843,0.001150455,0.0000637815,0.0002132597,0.0001430034],"genre_scores_gemma":[0.7448592,0.000008462862,0.2548136,0.00005260931,0.00004980453,0.000103931,0.00004574726,0.00001539668,0.00005124433],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8623455,"threshold_uncertainty_score":0.563772,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.190347442643677,"score_gpt":0.377449432259311,"score_spread":0.187101989615634,"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."}}