{"id":"W4413967892","doi":"10.1371/journal.pone.0329482","title":"Leveraging synthetic data produced from museum specimens to train adaptable species classification models","year":2025,"lang":"en","type":"article","venue":"PLoS ONE","topic":"Species Distribution and Climate Change","field":"Environmental Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Laurentian University; Canadian Museum of Nature; University of British Columbia","funders":"Mitacs","keywords":"Computer science; Pipeline (software); Artificial intelligence; Synthetic data; Bespoke; Modular design; Generalization; Domain (mathematical analysis); Adaptation (eye); Machine learning; Scalability; Pattern recognition (psychology); Database","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":[0.001328447,0.0008417319,0.0003763839,0.0009139517,0.0002701341,0.00067194,0.001173481,0.0008679798,0.001866302],"category_scores_gemma":[0.004564268,0.0004980289,0.001151646,0.0006669384,0.0005889848,0.0008663544,0.000957029,0.001112396,0.0007464841],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006098932,"about_ca_system_score_gemma":0.000567466,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004165868,"about_ca_topic_score_gemma":0.006483153,"domain_scores_codex":[0.9996305,0.00009812104,0.00002063558,0.0001478591,0.00007285258,0.00003007529],"domain_scores_gemma":[0.9982698,0.0008045351,0.0001183396,0.0004395621,0.0003110071,0.00005681317],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001457023,0.0001756988,0.01654592,0.0003712327,0.0001918054,0.0002462418,0.0002657075,0.7280225,0.03599631,0.002761625,0.007634379,0.2076429],"study_design_scores_gemma":[0.00001219392,0.00005508066,0.004000781,0.00002365581,0.00002098643,0.0001149896,0.00008407782,0.9775735,0.01102,0.002969242,0.004098576,0.0000269132],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3272806,0.0005461586,0.655239,0.00067247,0.0003162296,0.0003075184,0.003461818,0.007852248,0.004323999],"genre_scores_gemma":[0.6173326,0.0002452673,0.3717199,0.0002468849,0.00004366156,0.0003097006,0.007892685,0.0005918792,0.00161739],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.004165868,"threshold_uncertainty_score":0.008283198,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2737747122277578,"score_gpt":0.2663091799126477,"score_spread":0.007465532315110102,"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."}}