{"id":"W2808559829","doi":"10.1101/346809","title":"Machine learning to classify animal species in camera trap images: applications in ecology","year":2018,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Species Distribution and Climate Change","field":"Environmental Science","cited_by":60,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Saskatchewan","funders":"University of Georgia Research Foundation; Animal and Plant Health Inspection Service; University of Georgia; National Wildlife Research Center; Colorado Parks and Wildlife; U.S. Department of Agriculture; U.S. Department of Energy","keywords":"Artificial intelligence; Computer science; Laptop; Wildlife; Convolutional neural network; Camera trap; Sample (material); Machine learning; Computer vision; Cartography; Pattern recognition (psychology); Geography; Ecology","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","insufficient_payload"],"consensus_categories":["insufficient_payload"],"category_scores_codex":[0.0005000766,0.0004255815,0.0004721476,0.0002582549,0.0001475718,0.000121006,0.0006386472,0.0003972755,0.009835473],"category_scores_gemma":[0.0001429615,0.0004733414,0.00009095918,0.0009736835,0.0003007484,0.0001211406,0.0008992458,0.0008603005,0.001427639],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002028822,"about_ca_system_score_gemma":0.00007691408,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0005087079,"about_ca_topic_score_gemma":0.0007232368,"domain_scores_codex":[0.9973167,0.0001466487,0.0005207012,0.001028177,0.0003061305,0.0006816391],"domain_scores_gemma":[0.9987846,0.00005893243,0.0002140218,0.0006469205,0.00004764783,0.000247903],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.00003714555,0.0003066434,0.7566801,0.00005939022,0.00001428003,0.0000369261,0.00006003598,0.0001591728,0.2409702,0.0004127358,0.001261249,0.000002054941],"study_design_scores_gemma":[0.0003205124,0.00007088842,0.9395669,0.00004206864,0.00001246381,1.618947e-8,0.00006750828,0.0001555026,0.01385391,0.000001506972,0.04542565,0.0004830726],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9966865,0.0001327695,0.0001410694,0.0006808906,0.0002223273,0.0008958573,0.0003809847,0.0001645882,0.0006949925],"genre_scores_gemma":[0.998262,0.0001887109,0.0005097552,0.000248457,0.0001251069,0.000567155,0.000003434776,0.00005893604,0.00003638654],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.2271163,"threshold_uncertainty_score":0.9997718,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02035598408290201,"score_gpt":0.2371458418577547,"score_spread":0.2167898577748527,"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."}}