{"id":"W2001936727","doi":"10.1109/lsp.2007.905088","title":"Combining Gaussianized/Non-Gaussianized Features to Improve Speaker Diarization of Telephone Conversations","year":2007,"lang":"en","type":"article","venue":"IEEE Signal Processing Letters","topic":"Speech Recognition and Synthesis","field":"Computer Science","cited_by":27,"is_retracted":false,"has_abstract":true,"ca_institutions":"Computer Research Institute of Montréal","funders":"","keywords":"Speaker diarisation; Viterbi algorithm; Computer science; Cluster analysis; Speech recognition; Mixture model; Pattern recognition (psychology); Segmentation; Bayesian information criterion; Mel-frequency cepstrum; Feature (linguistics); Artificial intelligence; Speaker recognition; Word error rate; Test set; Hierarchical clustering; Feature extraction; 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":[0.0006241198,0.0002185642,0.0002877922,0.0003870081,0.0002169327,0.0002272898,0.0005059643,0.00009817316,0.00004578751],"category_scores_gemma":[0.00007015758,0.000208924,0.00009227268,0.0009615098,0.00008928416,0.0005574286,0.00005454964,0.0001961185,0.00005793147],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00007137498,"about_ca_system_score_gemma":0.0001027785,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00003823526,"about_ca_topic_score_gemma":0.000008248983,"domain_scores_codex":[0.9981204,0.00005462008,0.0004647182,0.0004599597,0.00048369,0.0004166166],"domain_scores_gemma":[0.9988885,0.0001930815,0.00027642,0.0002721489,0.0001800985,0.000189766],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.00004473612,0.00007033194,0.000160279,0.00004851119,0.00002719747,0.00002552959,0.001867588,0.00007026279,0.8362989,0.0001498837,0.001683875,0.1595529],"study_design_scores_gemma":[0.001241973,0.0000972802,0.007634538,0.0002310624,0.00003649126,0.00002974124,0.0002426536,0.008956688,0.9802839,0.0003383842,0.0004022305,0.0005051284],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.07256859,0.00002129329,0.9216519,0.003273856,0.0003855511,0.0002733495,0.000004008019,0.0001616685,0.001659765],"genre_scores_gemma":[0.912257,0.000001250142,0.08310059,0.004345713,0.0001476948,0.00001137971,0.00000515405,0.00002308455,0.0001081901],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8396884,"threshold_uncertainty_score":0.8519673,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01048864228849473,"score_gpt":0.2396504136676137,"score_spread":0.229161771379119,"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."}}