{"id":"W3217550460","doi":"10.1093/icesjms/fsab233","title":"Automatic discard registration in cluttered environments using deep learning and object tracking: class imbalance, occlusion, and a comparison to human review","year":2021,"lang":"en","type":"article","venue":"ICES Journal of Marine Science","topic":"Identification and Quantification in Food","field":"Biochemistry, Genetics and Molecular Biology","cited_by":22,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"European Maritime and Fisheries Fund; Ministry of Agriculture - Saskatchewan","keywords":"Artificial intelligence; Computer science; Computer vision; Object (grammar); Fish <Actinopterygii>; Pattern recognition (psychology); Tracking (education); Dorsum; Artificial neural network; Fishery; Anatomy; Biology","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.004375883,0.0004651138,0.0006715741,0.003267383,0.0003424067,0.001282974,0.0008542393,0.0008316776,0.001172716],"category_scores_gemma":[0.01164806,0.0001717315,0.0005304685,0.002176621,0.0004299769,0.001243986,0.0006724605,0.0002760905,0.0007236979],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005184947,"about_ca_system_score_gemma":0.0008291787,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002713413,"about_ca_topic_score_gemma":0.003046337,"domain_scores_codex":[0.9970379,0.0007053108,0.0003773155,0.0007493274,0.00102955,0.0001006069],"domain_scores_gemma":[0.9943093,0.001866521,0.0007956525,0.0007142645,0.002217531,0.00009669709],"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":[0.00035593,0.00005944656,0.01630312,0.001430703,0.0001999615,0.0003500571,0.0001892537,0.006358613,0.01488998,0.001387496,0.008101434,0.9503741],"study_design_scores_gemma":[0.0001228978,0.00118526,0.1026756,0.001645201,0.00146124,0.00562053,0.001398021,0.4686868,0.177212,0.009350065,0.2303271,0.0003152797],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3059162,0.07925472,0.5895879,0.001989674,0.001635174,0.0003320479,0.00217876,0.004590528,0.01451492],"genre_scores_gemma":[0.8260111,0.01858081,0.1407218,0.0006149686,0.000476031,0.0001316821,0.002980914,0.0004567122,0.01002606],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.004375883,"threshold_uncertainty_score":0.02314222,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02641112524738269,"score_gpt":0.33792697240222,"score_spread":0.3115158471548373,"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."}}