{"id":"W4364305273","doi":"10.1109/aipr57179.2022.10092213","title":"Achieving Adversarial Robustness in Deep Learning-Based Overhead Imaging","year":2022,"lang":"en","type":"article","venue":"","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Lockheed Martin (Canada)","funders":"","keywords":"Robustness (evolution); Computer science; Adversarial system; Artificial intelligence; Machine learning; Overhead (engineering); Deep learning; Object detection; Pipeline (software); Adversarial machine learning; Pattern recognition (psychology)","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":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.001425072,0.000266872,0.0003005948,0.0004824991,0.0007840613,0.0001847069,0.001708531,0.00004437583,0.0005893492],"category_scores_gemma":[0.0003426408,0.0003021608,0.0001233098,0.001283954,0.00005346327,0.0007766919,0.001651412,0.00140164,0.00001725172],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004449679,"about_ca_system_score_gemma":0.0001835325,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002385411,"about_ca_topic_score_gemma":0.00004835255,"domain_scores_codex":[0.9965934,0.0007763849,0.0004242637,0.0007841488,0.0007538354,0.0006680028],"domain_scores_gemma":[0.9986385,0.0004234894,0.0001965807,0.0005904846,0.00004498371,0.0001060028],"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":[0.00003103709,0.00006768288,0.033207,0.000008040452,0.000005295956,0.0001250229,0.0004174502,0.9451444,0.00007698671,0.00547853,0.00004318223,0.01539543],"study_design_scores_gemma":[0.001415481,0.00005731398,0.005295063,0.00001089578,0.000005127928,0.00002315515,0.0003787928,0.9901661,0.00002458878,0.0001656653,0.00208148,0.0003763168],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02154885,0.00008989808,0.9707043,0.001650869,0.00114508,0.0002133067,5.05057e-7,0.0005642056,0.004082992],"genre_scores_gemma":[0.9517787,0.000001004156,0.04707001,0.0005476873,0.0001173521,0.00005379613,0.000009264,0.00003410301,0.0003880948],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9302298,"threshold_uncertainty_score":0.9999431,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.006728922991036009,"score_gpt":0.2338659007135228,"score_spread":0.2271369777224868,"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."}}