{"id":"W6940722340","doi":"10.1093/icesjms/fsaf039","title":"Dataset selection is critical for effective pre-training of fish detection models for underwater video","year":2025,"lang":"en","type":"article","venue":"ICES Journal of Marine Science","topic":"Mycorrhizal Fungi and Plant Interactions","field":"Agricultural and Biological Sciences","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"National Research Council Canada; Fisheries and Oceans Canada; Community Sector Council Newfoundland and Labrador; Dalhousie University","funders":"National Research Council Canada; Natural Sciences and Engineering Research Council of Canada; Ocean Frontier Institute; Dalhousie University","keywords":"Underwater; Selection (genetic algorithm); Classifier (UML); Training set; Class (philosophy); Fish <Actinopterygii>; Marine fish; Training (meteorology); Data collection","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.006796344,0.002027285,0.001071008,0.001272545,0.001287428,0.002164768,0.002854365,0.00186328,0.004104981],"category_scores_gemma":[0.02167595,0.0007078272,0.00156909,0.001207676,0.0006681,0.002971811,0.001866448,0.003163577,0.004415807],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009945025,"about_ca_system_score_gemma":0.001953572,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01566498,"about_ca_topic_score_gemma":0.02634022,"domain_scores_codex":[0.9974293,0.0008745075,0.0002941869,0.0007893685,0.0003763993,0.0002361938],"domain_scores_gemma":[0.9921405,0.004412449,0.0003022547,0.001344105,0.001584029,0.000216704],"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":[0.001122641,0.001900478,0.1008034,0.001171487,0.0006268055,0.0005801572,0.0004423965,0.11964,0.03773067,0.002077008,0.1264532,0.6074516],"study_design_scores_gemma":[0.000265221,0.000582459,0.04198651,0.0006019899,0.0002959081,0.0003166464,0.000963776,0.817609,0.06067441,0.006770753,0.06976087,0.0001725373],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.4785661,0.003896919,0.4015597,0.004850359,0.002146435,0.002380646,0.04272527,0.04943942,0.01443527],"genre_scores_gemma":[0.5239934,0.0006525791,0.3493707,0.002028818,0.000184969,0.002031428,0.1155938,0.001498644,0.004645566],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01566498,"threshold_uncertainty_score":0.03594291,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0218799549263988,"score_gpt":0.2972444066898733,"score_spread":0.2753644517634745,"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."}}