{"id":"W4411757453","doi":"10.1145/3746638","title":"Latency-Aware Pruning and Quantization of Self-Supervised Speech Transformers for Edge Devices","year":2025,"lang":"en","type":"article","venue":"ACM Transactions on Embedded Computing Systems","topic":"Speech Recognition and Synthesis","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Computer science; Quantization (signal processing); Transformer; Speech recognition; Artificial intelligence; Algorithm; Engineering; Electrical engineering","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":[0.0004944508,0.0001963771,0.0003453509,0.0003972059,0.0003470598,0.0001475854,0.0004674539,0.0001297003,0.000003837954],"category_scores_gemma":[0.00003606639,0.0001940597,0.000127014,0.00067997,0.00003197166,0.0002552192,0.00001020949,0.0001255691,0.000003343621],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004817056,"about_ca_system_score_gemma":0.00009355381,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00004932297,"about_ca_topic_score_gemma":0.00001247899,"domain_scores_codex":[0.9984494,0.0001250791,0.0005236971,0.0004311344,0.0002130514,0.0002575711],"domain_scores_gemma":[0.9985145,0.0006735845,0.0001392583,0.0003667812,0.0002389265,0.00006697366],"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":[0.00005631519,0.0003739472,0.0003572983,0.001748577,0.0004383438,0.000003824307,0.005727357,0.002988337,0.002224467,0.003947496,0.0001060943,0.9820279],"study_design_scores_gemma":[0.001041543,0.0001311985,0.0002183877,0.0006905143,0.00008648813,0.00002163789,0.002044598,0.9824082,0.01238201,0.0002746949,0.0004364763,0.0002642994],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.04525533,0.0001261989,0.9518867,0.0003065252,0.000740231,0.0006710518,0.00001066705,0.0003327387,0.0006705616],"genre_scores_gemma":[0.8950467,0.00003148262,0.1046698,0.00009374342,0.00002232567,0.00003018926,0.000004335951,0.00001331517,0.00008817502],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9817637,"threshold_uncertainty_score":0.7913525,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02292530309430112,"score_gpt":0.2754509292688123,"score_spread":0.2525256261745112,"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."}}