{"id":"W2891182582","doi":"10.1016/j.ecoinf.2018.09.007","title":"A Deep learning method for accurate and fast identification of coral reef fishes in underwater images","year":2018,"lang":"en","type":"article","venue":"Ecological Informatics","topic":"Ichthyology and Marine Biology","field":"Environmental Science","cited_by":244,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Toronto","funders":"Centre Méditerranéen de l’Environnement et de la Biodiversité","keywords":"Identification (biology); Convolutional neural network; Underwater; Computer science; Artificial intelligence; Coral reef fish; Task (project management); Coral reef; Fish <Actinopterygii>; Deep learning; Pattern recognition (psychology); Fishery; Machine learning; Ecology; Biology; Geology; Oceanography","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":[0.0005235009,0.00005747531,0.0001186832,0.0000219721,0.00006701623,0.00000970901,0.0000862029,0.0001022937,0.0003326968],"category_scores_gemma":[0.0002331247,0.00004153309,0.00001652709,0.00005690489,0.0002903013,0.0001592582,0.0001611376,0.00007785897,0.00007003224],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002152594,"about_ca_system_score_gemma":0.000002206861,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001592737,"about_ca_topic_score_gemma":0.0001519185,"domain_scores_codex":[0.9993871,0.00005616187,0.0003098887,0.00006784419,0.00003382263,0.0001451112],"domain_scores_gemma":[0.9996184,0.0001735355,0.0001187153,0.00005685689,0.00001122624,0.00002125838],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0004797798,0.000377324,0.8052076,0.0001683956,0.00005312659,0.000003511203,0.01088482,0.005454935,0.0261753,0.002883743,0.00274249,0.1455691],"study_design_scores_gemma":[0.0003796187,0.0006059765,0.8861356,0.000002326417,0.000009268799,0.000007809326,0.0006560507,0.102796,0.003101213,0.004221318,0.001984641,0.0001002346],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9152464,0.000002684733,0.08104917,0.0001569488,0.0000630647,0.0001994974,0.000001973693,0.00001391089,0.003266319],"genre_scores_gemma":[0.9695752,0.000009494323,0.02985942,0.0002089862,0.0000109372,0.00001782325,0.00001162939,0.000001819937,0.0003046851],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.1454688,"threshold_uncertainty_score":0.3642797,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01922493413637658,"score_gpt":0.2891331268245805,"score_spread":0.2699081926882039,"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."}}