{"id":"W4381194140","doi":"10.1016/j.infsof.2023.107281","title":"Robustness assessment of hyperspectral image CNNs using metamorphic testing","year":2023,"lang":"en","type":"article","venue":"Information and Software Technology","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":5,"is_retracted":false,"has_abstract":false,"ca_institutions":"Polytechnique Montréal","funders":"","keywords":"Robustness (evolution); Computer science; Convolutional neural network; Artificial intelligence; Hyperspectral imaging; Machine learning; Deep learning; Leverage (statistics); 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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00204987,0.0007554783,0.0004083768,0.0008322559,0.0002519904,0.0007669923,0.0009070297,0.0009604096,0.001350021],"category_scores_gemma":[0.009287259,0.0002464301,0.0005537402,0.0003641348,0.0009028952,0.0009690436,0.001114544,0.0007392153,0.0001891865],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008524371,"about_ca_system_score_gemma":0.0005080909,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002355699,"about_ca_topic_score_gemma":0.001858289,"domain_scores_codex":[0.9992899,0.0001638431,0.00003644659,0.0001426353,0.0002827054,0.00008449216],"domain_scores_gemma":[0.99698,0.001512635,0.0003715144,0.0004715903,0.0005599069,0.0001043686],"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":[0.0005520771,0.00007783675,0.004874304,0.0001193556,0.0001830058,0.0001636046,0.00004437869,0.8902702,0.02888356,0.006694729,0.0009513219,0.06718566],"study_design_scores_gemma":[0.000003648592,0.00005006125,0.001118141,0.000007638537,0.00001146376,0.00003892095,0.000005917145,0.9896535,0.008005418,0.000982361,0.000117239,0.000005671133],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.4958724,0.0006687134,0.4946565,0.0005743419,0.0001378774,0.0001030087,0.0003106313,0.001205571,0.006470967],"genre_scores_gemma":[0.9771014,0.00009029731,0.0217258,0.00006014719,0.00001405142,0.00001983066,0.0001847223,0.00005358762,0.0007501457],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.002355699,"threshold_uncertainty_score":0.01084083,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02552026359889529,"score_gpt":0.2976550759759764,"score_spread":0.2721348123770811,"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."}}