{"id":"W4283702278","doi":"10.11591/ijai.v11.i3.pp961-968","title":"Defense against adversarial attacks on deep convolutional neural networks through nonlocal denoising","year":2022,"lang":"en","type":"article","venue":"IAES International Journal of Artificial Intelligence","topic":"Image and Signal Denoising Methods","field":"Computer Science","cited_by":10,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Universiti Brunei Darussalam","keywords":"MNIST database; Adversarial system; Computer science; Artificial intelligence; Convolutional neural network; Deep learning; Deep neural networks; Robustness (evolution); Machine learning; Classifier (UML); Artificial neural network; Luminance","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001167288,0.0006610189,0.0005006653,0.000386135,0.00026655,0.0004561557,0.0006857034,0.0007635003,0.000974172],"category_scores_gemma":[0.002749188,0.000236809,0.000427976,0.0001915914,0.00111463,0.001012481,0.001548874,0.001611017,0.000297193],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005461242,"about_ca_system_score_gemma":0.0003970719,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0009340512,"about_ca_topic_score_gemma":0.001257094,"domain_scores_codex":[0.9994942,0.0001427554,0.00001777442,0.00008590676,0.0001858418,0.00007355454],"domain_scores_gemma":[0.9989172,0.0005303374,0.0001288843,0.0002484746,0.0001305992,0.00004448031],"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":[0.0002326288,0.00006294766,0.00106828,0.00006161145,0.00008836096,0.0001676337,0.00008997801,0.8361223,0.04894506,0.02984028,0.001967337,0.08135364],"study_design_scores_gemma":[0.00000437991,0.00003690134,0.000141231,0.000007348215,0.000004790064,0.00002898838,0.000006554201,0.984021,0.01034745,0.004945435,0.0004511304,0.000004696215],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.08693627,0.0003572313,0.9067948,0.0005465754,0.00007036858,0.00002785788,0.00005255026,0.001048387,0.004166],"genre_scores_gemma":[0.926052,0.0002916661,0.06934484,0.0002446962,0.0000388484,0.0000363572,0.00009887211,0.00009577157,0.003796847],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.001167288,"threshold_uncertainty_score":0.006173253,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04769861058786611,"score_gpt":0.3305586075210692,"score_spread":0.2828599969332031,"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."}}