{"id":"W2780788572","doi":"10.3390/jimaging5020026","title":"PixelBNN: Augmenting the PixelCNN with Batch Normalization and the Presentation of a Fast Architecture for Retinal Vessel Segmentation","year":2019,"lang":"en","type":"article","venue":"Journal of Imaging","topic":"Retinal Imaging and Analysis","field":"Medicine","cited_by":54,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Segmentation; Normalization (sociology); Fundus (uterus); Image segmentation; Retinal; Hue; Image quality; Image processing","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":[0.0007946722,0.00008032988,0.0002129467,0.00009938777,0.00008875389,0.00005414688,0.00006504916,0.00001042565,0.000007965427],"category_scores_gemma":[0.00009354536,0.00003638518,0.00009188028,0.0001642446,0.0001011179,0.0001603795,0.00001424463,0.0001487113,2.867612e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001748633,"about_ca_system_score_gemma":0.00004181078,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002165415,"about_ca_topic_score_gemma":0.000001333038,"domain_scores_codex":[0.9991154,0.00007818396,0.0003263044,0.0000800549,0.0002997217,0.0001003561],"domain_scores_gemma":[0.9988001,0.0001672199,0.0005802848,0.0000961046,0.0003282437,0.00002807709],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.005183606,0.0001233242,0.7207936,0.0008123362,0.0007436099,0.0000132231,0.015959,0.00412223,0.1690857,0.0003000911,0.0009002228,0.08196305],"study_design_scores_gemma":[0.06062379,0.00229174,0.2881522,0.00734696,0.01010197,0.007715612,0.09324792,0.4016239,0.1214725,0.00255019,0.004039135,0.000834047],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8217707,0.0004337874,0.1661065,0.01095883,0.00005775048,0.0003429547,0.000001226653,0.000004543847,0.0003237434],"genre_scores_gemma":[0.9940866,0.00004607073,0.005324466,0.0002319437,0.0001311018,0.000003689877,0.000005189145,0.00001070647,0.0001601803],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.4326414,"threshold_uncertainty_score":0.1483744,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.004571420200184811,"score_gpt":0.2578317520577964,"score_spread":0.2532603318576116,"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."}}