{"id":"W4295308567","doi":"10.1109/jstsp.2022.3206084","title":"Self-Supervised Language Learning From Raw Audio: Lessons From the Zero Resource Speech Challenge","year":2022,"lang":"en","type":"article","venue":"IEEE Journal of Selected Topics in Signal Processing","topic":"Speech Recognition and Synthesis","field":"Computer Science","cited_by":26,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"Natural Sciences and Engineering Research Council of Canada; University of Toronto; Connaught Fund; Agence Nationale de la Recherche; École des Hautes Etudes en Sciences Sociales","keywords":"Computer science; Speech recognition; Zero (linguistics); Resource (disambiguation); Artificial intelligence; Natural language processing; Linguistics","routes":{"ca_aff":true,"ca_fund":true,"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":[0.01723152,0.001112932,0.001837911,0.001088362,0.0008750057,0.003406656,0.002914007,0.002566764,0.001862834],"category_scores_gemma":[0.04081148,0.0004660038,0.0007150406,0.001151401,0.003402576,0.006867487,0.003671241,0.006668616,0.0021848],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001167994,"about_ca_system_score_gemma":0.002063931,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004151074,"about_ca_topic_score_gemma":0.005206897,"domain_scores_codex":[0.9934071,0.003434058,0.0003189661,0.001240587,0.001424066,0.0001752835],"domain_scores_gemma":[0.9500929,0.04012795,0.0004440281,0.003665848,0.004770803,0.000898474],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0004850576,0.0003039036,0.003946382,0.001240063,0.0002320065,0.0002662943,0.00108021,0.07520425,0.002788536,0.08698076,0.1189323,0.7085403],"study_design_scores_gemma":[0.0001018379,0.0002442306,0.002225201,0.0003887546,0.00004153153,0.0003461514,0.0008544107,0.4605196,0.006119276,0.4409166,0.08809896,0.000143431],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.03442206,0.03891567,0.8439711,0.05810812,0.002937589,0.0001569667,0.002428579,0.003317695,0.01574221],"genre_scores_gemma":[0.3991232,0.02369288,0.5305943,0.009733899,0.008100796,0.0005127578,0.01155053,0.002058157,0.01463356],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01723152,"threshold_uncertainty_score":0.09113002,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03068583258178572,"score_gpt":0.263617653528441,"score_spread":0.2329318209466552,"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."}}