{"id":"W2922342173","doi":"10.1109/tip.2019.2904267","title":"Shadow Detection in Single RGB Images Using a Context Preserver Convolutional Neural Network Trained by Multiple Adversarial Examples","year":2019,"lang":"en","type":"article","venue":"IEEE Transactions on Image Processing","topic":"Digital Media Forensic Detection","field":"Computer Science","cited_by":35,"is_retracted":false,"has_abstract":true,"ca_institutions":"Simon Fraser University","funders":"Stony Brook University; Simon Fraser University; Government of Canada","keywords":"Artificial intelligence; Convolutional neural network; Computer science; Context (archaeology); Adversarial system; RGB color model; Computer vision; Pattern recognition (psychology); Shadow (psychology); Artificial neural network; Image processing; Contextual image classification; Image (mathematics)","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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0002618406,0.0003007649,0.0002943258,0.0002569423,0.0002697735,0.0005057828,0.0003815248,0.000135864,0.00002628291],"category_scores_gemma":[0.00004049801,0.0003240274,0.0001281128,0.0008382834,0.0001469383,0.003656704,0.00000959716,0.0004002318,0.00003914534],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00030759,"about_ca_system_score_gemma":0.0001211824,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003014266,"about_ca_topic_score_gemma":0.0003874122,"domain_scores_codex":[0.9977135,0.0001156081,0.0004506943,0.0006921294,0.0004566703,0.0005714207],"domain_scores_gemma":[0.9990278,0.0002232477,0.0001416182,0.0003223519,0.0001648504,0.0001201016],"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":[0.0003606137,0.0004550635,0.0001546227,0.0001054239,0.00003440549,0.00001126596,0.0006857392,0.05522643,0.3741124,0.000009475547,0.00009370184,0.5687509],"study_design_scores_gemma":[0.002298909,0.0002662228,0.0003367459,0.0001748767,0.00002030572,0.00005937184,0.0001336309,0.8132848,0.1824491,0.000355166,0.0001808447,0.000440028],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2362677,0.00009457099,0.7612194,0.00007796427,0.001489745,0.0004109863,0.00001877349,0.0002426023,0.0001781419],"genre_scores_gemma":[0.9824905,0.000001907642,0.01706726,0.0001168103,0.0001179027,0.00004578388,0.000004801238,0.00003591671,0.0001191465],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.7580584,"threshold_uncertainty_score":0.9999212,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01891916739155991,"score_gpt":0.2269450545083561,"score_spread":0.2080258871167962,"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."}}