{"id":"W4298394324","doi":"10.1145/3552466.3556523","title":"A Comparative Study on Physical and Perceptual Features for Deepfake Audio Detection","year":2022,"lang":"en","type":"article","venue":"","topic":"Music and Audio Processing","field":"Computer Science","cited_by":32,"is_retracted":false,"has_abstract":true,"ca_institutions":"Toronto Metropolitan University","funders":"","keywords":"Computer science; Perception; Feature (linguistics); Artificial intelligence; Feature extraction; Speech recognition; Machine learning","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":[0.0001024421,0.00007378127,0.0001129836,0.00003881509,0.0004831163,0.00008723115,0.0001590701,0.000007892178,0.000007379236],"category_scores_gemma":[0.00000528122,0.00005997286,0.00002550749,0.0001286074,0.00001932203,0.0001241895,0.0001603309,0.0001048628,0.000002020995],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0000254041,"about_ca_system_score_gemma":0.00001581803,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000007176216,"about_ca_topic_score_gemma":0.00001719768,"domain_scores_codex":[0.9993482,0.00005188197,0.00005614507,0.0002676263,0.0001658803,0.0001103287],"domain_scores_gemma":[0.9997358,0.00007475851,0.00002973723,0.0001105529,0.00001989553,0.00002926802],"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":[0.0003737729,0.003628986,0.0008013367,0.00004168725,0.000158927,0.0000170056,0.388861,0.001934667,0.03235845,0.0287998,0.0160585,0.5269659],"study_design_scores_gemma":[0.008660586,0.02341829,0.2202217,0.00002615593,0.0001093584,0.0001395815,0.1177907,0.5484514,0.04837469,0.01576037,0.01515814,0.001889054],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7986624,0.0000114175,0.1995683,0.0002347854,0.0001018549,0.000270694,9.61734e-7,0.00007092286,0.001078729],"genre_scores_gemma":[0.9973635,8.211751e-8,0.00163304,0.0004614124,0.00006727005,0.000096494,4.825405e-7,0.000002959367,0.0003747774],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.5465167,"threshold_uncertainty_score":0.3715788,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03890443042575863,"score_gpt":0.3082676728180773,"score_spread":0.2693632423923186,"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."}}