{"id":"W7125949704","doi":"10.1109/smc58881.2025.11343334","title":"Phonetic Analysis of Real and Synthetic Speech Using HuBERT Embeddings: Perspectives for Deepfake Detection","year":2025,"lang":"en","type":"article","venue":"","topic":"Speech Recognition and Synthesis","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"Institut National de la Recherche Scientifique; Cégep de l'Outaouais","funders":"","keywords":"Speech synthesis; Synthetic data; Sophistication; Speech processing; Rank (graph theory); Divergence (linguistics); Noise (video)","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":[0.0006812951,0.0009326977,0.0005488343,0.001072853,0.0002433488,0.001024734,0.0005009114,0.0007987147,0.001317504],"category_scores_gemma":[0.002392946,0.0002337794,0.0004299335,0.0006153689,0.000540686,0.001297755,0.0007445528,0.001096836,0.001126472],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004193242,"about_ca_system_score_gemma":0.0004980516,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002814301,"about_ca_topic_score_gemma":0.004120388,"domain_scores_codex":[0.9996184,0.0001001668,0.0000236895,0.0001180411,0.00007638245,0.00006324301],"domain_scores_gemma":[0.9990624,0.0004058588,0.000094736,0.0001618468,0.0002090065,0.00006625122],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0007524719,0.00037488,0.01306118,0.0002975566,0.0001512401,0.0005026431,0.0004977616,0.2328652,0.1077152,0.01011897,0.005038647,0.6286243],"study_design_scores_gemma":[0.000009715015,0.0001177868,0.005497652,0.00002734105,0.0000205796,0.0001298291,0.0001917521,0.961518,0.02384049,0.006003538,0.002609959,0.00003338558],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4011805,0.002171736,0.5867274,0.0008889935,0.0002346783,0.00008897225,0.001458065,0.003099967,0.004149771],"genre_scores_gemma":[0.8805142,0.0005505754,0.1108093,0.000136968,0.0000885914,0.00006535318,0.003019928,0.0002426709,0.004572366],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.002814301,"threshold_uncertainty_score":0.005595803,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01738696341447149,"score_gpt":0.2900737172972532,"score_spread":0.2726867538827817,"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."}}