{"id":"W4383682007","doi":"10.35490/ec3.2023.321","title":"Enhancing single-stage excavator activity recognition via knowledge distillation of temporal gradient data","year":2023,"lang":"en","type":"article","venue":"Computing in construction","topic":"Occupational Health and Safety Research","field":"Health Professions","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Concordia University","funders":"","keywords":"Computer science; Excavator; Distillation; Frame (networking); Construct (python library); Artificial intelligence; Activity recognition; Exploit; Pattern recognition (psychology); Domain knowledge; Data mining; Machine learning; Computer vision; 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":[],"consensus_categories":[],"category_scores_codex":[0.002140794,0.0001115585,0.000238961,0.0003707188,0.0004681177,0.000006426308,0.000159182,0.0001537273,0.00006191251],"category_scores_gemma":[0.0008897784,0.0001193363,0.00002874057,0.0009835432,0.00009572117,0.0002237518,0.0002767882,0.000476269,0.000147821],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000304678,"about_ca_system_score_gemma":0.0003931286,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0004962012,"about_ca_topic_score_gemma":0.0008248103,"domain_scores_codex":[0.9976142,0.0006444305,0.0006843825,0.0003762638,0.0002551491,0.0004255581],"domain_scores_gemma":[0.997367,0.001534511,0.0003729844,0.0003638583,0.0002620721,0.00009958793],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.0002114279,0.00009404233,0.4811389,0.00119815,0.0000094962,0.000002385319,0.000891225,0.0001210717,0.003514268,0.0001376371,0.0001455785,0.5125358],"study_design_scores_gemma":[0.0009258526,0.00009885591,0.7859063,0.0008024462,0.000006989213,0.000003403847,0.0007859648,0.2090297,0.0007558856,0.0006183319,0.0008922672,0.0001740554],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9733917,0.00002510737,0.02256906,0.0001274225,0.001866153,0.0007246621,0.0001066922,0.0001373984,0.001051767],"genre_scores_gemma":[0.9960521,0.00001506266,0.002804165,0.000008761455,0.0003673999,0.00001535328,0.0006656169,0.00001591824,0.0000556384],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.5123618,"threshold_uncertainty_score":0.4866394,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2533656659201612,"score_gpt":0.4799525884914176,"score_spread":0.2265869225712564,"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."}}