{"id":"W4385301593","doi":"10.1016/j.ocecoaman.2023.106765","title":"MLDet: Towards efficient and accurate deep learning method for Marine Litter Detection","year":2023,"lang":"en","type":"article","venue":"Ocean & Coastal Management","topic":"Microplastics and Plastic Pollution","field":"Environmental Science","cited_by":34,"is_retracted":false,"has_abstract":false,"ca_institutions":"Université du Québec à Rimouski","funders":"National Natural Science Foundation of China","keywords":"Marine debris; Litter; Underwater; Environmental science; Computer science; Artificial intelligence; Oceanography; Ecology; Biology; Geology","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":[0.0004198714,0.0001279396,0.00009750851,0.00007136923,0.0002010152,0.00005853866,0.00007786912,0.00003240382,0.0002355681],"category_scores_gemma":[0.00002916862,0.0001192956,0.00003830279,0.0002313669,0.00003705326,0.00004504243,0.0005646333,0.00007623449,0.000173966],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005890166,"about_ca_system_score_gemma":0.000001345978,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001141012,"about_ca_topic_score_gemma":0.000131051,"domain_scores_codex":[0.9990276,0.00003714215,0.000149569,0.0003141247,0.0001778635,0.0002937127],"domain_scores_gemma":[0.9997027,0.00006180285,0.0000566526,0.0001005601,0.000005625559,0.00007267862],"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":[0.0001079635,0.00004665827,0.00104452,0.0001081676,0.00004823711,0.00002054659,0.0003185356,0.0975633,0.004091622,0.0004486916,0.001233992,0.8949677],"study_design_scores_gemma":[0.0007882726,0.0002255042,0.1165817,0.00001821896,0.00007751935,0.000008311001,0.0003885682,0.8007718,0.001119068,0.0007362312,0.0789873,0.0002975724],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2806214,0.000005755331,0.7093617,0.0001932723,0.000325857,0.0005131835,0.00002109866,0.0001159385,0.008841781],"genre_scores_gemma":[0.988929,0.00003727257,0.007190634,0.0000735028,0.0000399429,0.00001637459,0.00003830787,0.00001859459,0.00365636],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8946702,"threshold_uncertainty_score":0.4864733,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.00886209475991615,"score_gpt":0.2345428831304956,"score_spread":0.2256807883705794,"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."}}