{"id":"W3097523643","doi":"10.3390/app10217522","title":"Robust Deep Speaker Recognition: Learning Latent Representation with Joint Angular Margin Loss","year":2020,"lang":"en","type":"article","venue":"Applied Sciences","topic":"Speech Recognition and Synthesis","field":"Computer Science","cited_by":13,"is_retracted":false,"has_abstract":true,"ca_institutions":"Concordia University","funders":"","keywords":"Computer science; Softmax function; Speaker recognition; Speech recognition; Artificial intelligence; Biometrics; TIMIT; Pattern recognition (psychology); Codebook; Machine learning; Convolutional neural network; Hidden Markov model","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":[0.0004086276,0.0001376574,0.0001705256,0.00009212114,0.0003655252,0.0003672989,0.0004614161,0.00004024891,0.0003251637],"category_scores_gemma":[0.00006167837,0.0001080502,0.00004844875,0.001087032,0.0002067616,0.000443331,0.0001161493,0.0001453408,0.0005082993],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001846052,"about_ca_system_score_gemma":0.00005093663,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0000111488,"about_ca_topic_score_gemma":0.000008286309,"domain_scores_codex":[0.9982393,0.00006882338,0.0002111182,0.000640508,0.000558377,0.0002819129],"domain_scores_gemma":[0.9993998,0.00008364987,0.0001175236,0.0001672975,0.00007259301,0.0001591671],"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":[0.00009231982,0.0002259355,0.006850819,0.00005984329,0.00008864399,0.0002929581,0.007937954,0.01496099,0.01337708,0.01590019,0.00154199,0.9386713],"study_design_scores_gemma":[0.003763789,0.001728848,0.02611076,0.0002195028,0.0001204856,0.0004479056,0.01019061,0.6023045,0.3181684,0.02262629,0.01116679,0.003152159],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.213821,0.00006092313,0.7178828,0.008519362,0.0001656767,0.0004640423,0.00000168863,0.0005542259,0.05853032],"genre_scores_gemma":[0.8296731,0.00001845646,0.1687568,0.001331127,0.0001078664,0.00003357844,0.000005423018,0.000008214895,0.00006543181],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9355191,"threshold_uncertainty_score":0.6533329,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1056419227313543,"score_gpt":0.2373883588161174,"score_spread":0.1317464360847631,"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."}}