{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0008284162,0.0001757933,0.0003198725,0.0001818732,0.0005121863,0.0002684626,0.001236469,0.00008140576,0.0001815183],"category_scores_gemma":[0.0001557958,0.0001430546,0.0001054813,0.0008692889,0.00003076018,0.0003977858,0.000156874,0.001469638,0.000005285192],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001778815,"about_ca_system_score_gemma":0.0003455442,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00007655907,"about_ca_topic_score_gemma":0.00003673135,"domain_scores_codex":[0.9973969,0.0006440242,0.0005677892,0.0002920066,0.0007913809,0.0003078919],"domain_scores_gemma":[0.9983804,0.0006184356,0.0004932852,0.0001998546,0.0002164316,0.00009165846],"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":[0.00008554436,0.0003628676,0.001121176,0.00002307348,0.0001233575,0.0007121284,0.02281911,0.001725334,0.01835804,0.00008158082,0.0009342124,0.9536536],"study_design_scores_gemma":[0.008175573,0.001433823,0.01129236,0.001329508,0.000342758,0.001107535,0.01959295,0.7535428,0.1247945,0.02212555,0.05401676,0.002245893],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8462602,0.006356912,0.1281237,0.01565722,0.0006755281,0.0002473403,0.00000862108,0.0002250929,0.00244535],"genre_scores_gemma":[0.9723545,0.00007081447,0.02613378,0.0006581428,0.000619373,0.000006101849,0.00000460101,0.00002075876,0.0001319192],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9514077,"threshold_uncertainty_score":0.6384929,"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."}}