{"id":"W4224917505","doi":"10.1109/icassp43922.2022.9747452","title":"Robust Self-Supervised Speaker Representation Learning Via Instance Mix Regularization","year":2022,"lang":"en","type":"article","venue":"ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","topic":"Speech Recognition and Synthesis","field":"Computer Science","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"Computer Research Institute of Montréal","funders":"","keywords":"Computer science; Embedding; Regularization (linguistics); Speech recognition; Feature learning; Utterance; Artificial intelligence; Supervised learning; Speaker recognition; Pattern recognition (psychology); Machine learning; Artificial neural network","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","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0009147828,0.0004023032,0.0003885058,0.0005336237,0.001075418,0.0009464646,0.001170788,0.0001342936,0.002364568],"category_scores_gemma":[0.0002185361,0.0004432135,0.0001318015,0.0008233876,0.0001175784,0.0009170176,0.0004273545,0.001028704,0.00006444933],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003803858,"about_ca_system_score_gemma":0.0003169463,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00003034248,"about_ca_topic_score_gemma":0.000006872766,"domain_scores_codex":[0.9953347,0.0004199231,0.0007155594,0.001168109,0.001868776,0.0004928759],"domain_scores_gemma":[0.9979287,0.0002497176,0.0005216041,0.0003944318,0.0006693169,0.0002362187],"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":[0.0003984969,0.001106963,0.001152486,0.0001294681,0.0002618194,0.0006963394,0.002240188,0.01367304,0.1261531,0.01041644,0.004286848,0.8394848],"study_design_scores_gemma":[0.0008424856,0.0002749235,0.0003683492,0.00007607442,0.00004340411,0.0002427445,0.0009501852,0.9853699,0.0035989,0.005650127,0.002012044,0.0005708759],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02958039,0.0001596583,0.9476438,0.003961196,0.001565707,0.0005288467,0.00008191281,0.0006430764,0.01583542],"genre_scores_gemma":[0.9340276,0.0001924535,0.0602501,0.001054383,0.0003657485,0.0001298067,0.0001731299,0.00005411437,0.003752665],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9716969,"threshold_uncertainty_score":0.9998019,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05941694244909695,"score_gpt":0.2801079894117903,"score_spread":0.2206910469626934,"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."}}