{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,11,18]],"date-time":"2026-11-18T08:19:53Z","timestamp":1794989993292,"version":"4.9.1"},"reference-count":344,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","license":[{"start":{"date-parts":[[2024,1,1]],"date-time":"2024-01-01T00:00:00Z","timestamp":1704067200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/legalcode"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE\/ACM Trans. Audio Speech Lang. Process."],"published-print":{"date-parts":[[2024]]},"DOI":"10.1109\/taslp.2023.3328283","type":"journal-article","created":{"date-parts":[[2023,10,30]],"date-time":"2023-10-30T15:26:29Z","timestamp":1698679589000},"page":"325-351","source":"Crossref","is-referenced-by-count":226,"title":["End-to-End Speech Recognition: A Survey"],"prefix":"10.1109","volume":"32","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-5331-6058","authenticated-orcid":false,"given":"Rohit","family":"Prabhavalkar","sequence":"first","affiliation":[{"name":"Google LLC., Mountain View, CA, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4560-8039","authenticated-orcid":false,"given":"Takaaki","family":"Hori","sequence":"additional","affiliation":[{"name":"Apple Inc., Cambridge, MA, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4126-6556","authenticated-orcid":false,"given":"Tara N.","family":"Sainath","sequence":"additional","affiliation":[{"name":"Google LLC., New York, NY, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2839-9247","authenticated-orcid":false,"given":"Ralf","family":"Schl\u00fcter","sequence":"additional","affiliation":[{"name":"Lehrstuhl Informatik 6 - Computer Science Department, RWTH Aachen University, Aachen, Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5970-8631","authenticated-orcid":false,"given":"Shinji","family":"Watanabe","sequence":"additional","affiliation":[{"name":"Carnegie Mellon University, Pittsburgh, PA, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","first-page":"370","article-title":"An essay towards solving a problem in the doctrine of chances","volume":"53","author":"Bayes","year":"1763","journal-title":"Philos. Trans. Roy. Soc. London"},{"key":"ref2","volume-title":"Statistical Methods for Speech Recognition","author":"Jelinek","year":"1997"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1109\/5.18626"},{"key":"ref4","volume-title":"Connectionist Speech Recognition: A Hybrid Approach","author":"Bourlard","year":"1993"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2011-169"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.21437\/Eurospeech.1997-548"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2000.862024"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1109\/icassp.1989.266531"},{"key":"ref9","first-page":"932","article-title":"A neural probabilistic language model","volume-title":"Proc. Neural Inf. Process. Syst.","author":"Bengio","year":"2000"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2002.5743830"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.21437\/interspeech.2014-223"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1109\/ASRU.2015.7404770"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1145\/1143844.1143891"},{"key":"ref14","article-title":"Sequence transduction with recurrent neural networks","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Graves","year":"2012"},{"key":"ref15","first-page":"577","article-title":"Attention-based models for speech recognition","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Chorowski","year":"2015"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2016.7472621"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.3115\/1220175.1220271"},{"key":"ref18","first-page":"2493","article-title":"Natural language processing (Almost) from scratch","volume":"12","author":"Collobert","year":"2011","journal-title":"J. Mach. Learn. Res."},{"key":"ref19","first-page":"1764","article-title":"Towards end-to-end speech recognition with recurrent neural networks","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Graves","year":"2014"},{"key":"ref20","article-title":"Cambridge Dictionary","year":"2020"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2018-1025"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2019.8682336"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.1109\/97.917693"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-540-44871-6_74"},{"key":"ref25","first-page":"12449","article-title":"wav2vec 2.0: A framework for self-supervised learning of speech representations","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Baevski","year":"2020"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1109\/ASRU46091.2019.9003986"},{"key":"ref27","volume-title":"Classication and Regression Trees","author":"Breiman","year":"1984"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1109\/ASRU.2017.8268950"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2019-3167"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2017-1566"},{"key":"ref31","volume-title":"Human Behavior and the Principle of Least Effort","author":"Zipf","year":"1949"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/p16-1162"},{"key":"ref33","article-title":"Latent sequence decompositions","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Chan","year":"2017"},{"key":"ref34","first-page":"2188","article-title":"Gram-CTC: Automatic unit selection and target decomposition for sequence labelling","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Liu","year":"2017"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2019.8682494"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2021-1623"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2012.6289079"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1006\/csla.2001.0184"},{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2018-1212"},{"key":"ref40","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2021-1671"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2017-233"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.1162\/neco.1997.9.8.1735"},{"key":"ref43","doi-asserted-by":"publisher","DOI":"10.1146\/annurev.neuro.26.041002.131047"},{"key":"ref44","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.1706.03762"},{"key":"ref45","doi-asserted-by":"publisher","DOI":"10.21437\/interspeech.2020-3015"},{"key":"ref46","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2017-1705"},{"key":"ref47","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP40776.2020.9053600"},{"key":"ref48","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2013.6638947"},{"key":"ref49","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP43922.2022.9747788"},{"key":"ref50","article-title":"Estimating or propagating gradients through stochastic neurons for conditional computation","author":"Bengio","year":"2013"},{"key":"ref51","doi-asserted-by":"publisher","DOI":"10.1109\/ASRU46091.2019.9003822"},{"key":"ref52","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2020-1855"},{"key":"ref53","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2016.7472618"},{"key":"ref54","article-title":"Googles neural machine translation system: Bridging the gap between human and machine translation","author":"Wu","year":"2016"},{"key":"ref55","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2018-1160"},{"key":"ref56","doi-asserted-by":"publisher","DOI":"10.31390\/gradschool_dissertations.4601"},{"key":"ref57","doi-asserted-by":"publisher","DOI":"10.21437\/interspeech.2017-751"},{"key":"ref58","doi-asserted-by":"publisher","DOI":"10.1109\/ASRU46091.2019.9003854"},{"key":"ref59","first-page":"5067","article-title":"An online sequence-to-sequence model using partial conditioning","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Jaitly","year":"2016"},{"key":"ref60","first-page":"2837","article-title":"Online and linear-time attention by enforcing monotonic alignments","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Raffel","year":"2017"},{"key":"ref61","article-title":"Monotonic Chunkwise attention","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Chiu","year":"2018"},{"key":"ref62","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P19-1126"},{"key":"ref63","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2018.8462366"},{"key":"ref64","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2019.8683510"},{"key":"ref65","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2019-2879"},{"key":"ref66","article-title":"A Study of Latent Monotonic Attention Variants","author":"Zeyer","year":"2021"},{"key":"ref67","doi-asserted-by":"publisher","DOI":"10.1109\/SLT54892.2023.10022818"},{"key":"ref68","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP40776.2020.9054260"},{"key":"ref69","doi-asserted-by":"publisher","DOI":"10.21437\/interspeech.2016-595"},{"key":"ref70","article-title":"Wav2Letter: An end-to-end convnet-based speech recognition system","author":"Collobert","year":"2016"},{"key":"ref71","doi-asserted-by":"publisher","DOI":"10.21437\/Eurospeech.1993-398"},{"key":"ref72","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2017-1073"},{"key":"ref73","doi-asserted-by":"publisher","DOI":"10.1109\/SLT54892.2023.10022967"},{"key":"ref74","doi-asserted-by":"publisher","DOI":"10.1109\/ASRU.2015.7404790"},{"key":"ref75","article-title":"Deep Speech: Scaling up End-to-End Speech Recognition","author":"Hannun","year":"2014"},{"key":"ref76","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2016.7472641"},{"key":"ref77","doi-asserted-by":"publisher","DOI":"10.21437\/interspeech.2017-343"},{"key":"ref78","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2017.7953077"},{"key":"ref79","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2017-1118"},{"key":"ref80","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2018-2456"},{"key":"ref81","article-title":"Optimal completion distillation for sequence learning","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Sabour","year":"2019"},{"key":"ref82","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2018-1030"},{"key":"ref83","doi-asserted-by":"publisher","DOI":"10.1109\/ASRU46091.2019.9003972"},{"key":"ref84","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2002.5743871"},{"key":"ref85","doi-asserted-by":"publisher","DOI":"10.1145\/3079856.3080246"},{"key":"ref86","doi-asserted-by":"publisher","DOI":"10.1109\/JSTSP.2017.2763455"},{"key":"ref87","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2019-1341"},{"key":"ref88","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP40776.2020.9053606"},{"key":"ref89","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP39728.2021.9414607"},{"key":"ref90","article-title":"Transformer transducer: One model unifying streaming and non-streaming speech recognition","author":"Tripathi","year":"2020"},{"key":"ref91","article-title":"Universal ASR: Unify and improve streaming asr with full-context modeling","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Yu","year":"2021"},{"key":"ref92","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2019-1209"},{"key":"ref93","doi-asserted-by":"publisher","DOI":"10.1109\/SLT.2018.8639034"},{"key":"ref94","doi-asserted-by":"publisher","DOI":"10.1109\/SLT.2018.8639044"},{"key":"ref95","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2019.8682441"},{"key":"ref96","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2018-1438"},{"key":"ref97","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2020-2059"},{"key":"ref98","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2018.8462506"},{"key":"ref99","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP40776.2020.9053896"},{"key":"ref100","first-page":"7829","article-title":"Transformer-transducer: End-to-snd speech recognition with self-attention","volume-title":"Proc. IEEE Int. Conf. Acoust., Speech, Signal Process.","author":"Yeh","year":"2019"},{"key":"ref101","first-page":"17627","article-title":"Branchformer: Parallel MLP-attention architectures to capture local and global context for speech recognition and understanding","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Peng","year":"2022"},{"key":"ref102","doi-asserted-by":"publisher","DOI":"10.1109\/ASRU46091.2019.9003750"},{"key":"ref103","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP39728.2021.9414858"},{"key":"ref104","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2021-212"},{"key":"ref105","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP40776.2020.9054419"},{"key":"ref106","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP39728.2021.9413648"},{"key":"ref107","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP39728.2021.9414212"},{"key":"ref108","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP43922.2022.9746908"},{"key":"ref109","doi-asserted-by":"publisher","DOI":"10.1109\/SLT54892.2023.10023194"},{"key":"ref110","article-title":"VioLA: Unified codec language models for speech recognition, synthesis, and translation","author":"Wang","year":"2023"},{"key":"ref111","article-title":"AudioPaLM: A large language model that can speak and listen","author":"Rubenstein","year":"2023"},{"key":"ref112","doi-asserted-by":"publisher","DOI":"10.1109\/ASRU46091.2019.9003787"},{"key":"ref113","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP40776.2020.9054715"},{"key":"ref114","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP40776.2020.9054358"},{"key":"ref115","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2021-1131"},{"key":"ref116","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-642-35289-8_26"},{"key":"ref117","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.1404.7828"},{"key":"ref118","first-page":"1","article-title":"An inequality and associated maximization technique in statistical estimation for probabilistic functions of Markov processes","volume":"3","author":"Baum","year":"1972","journal-title":"Inequalities"},{"key":"ref119","doi-asserted-by":"publisher","DOI":"10.1109\/MASSP.1986.1165342"},{"key":"ref120","first-page":"175","article-title":"Neural network-Gaussian mixture hybrid for speech recognition or density estimation","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Bengio","year":"1991"},{"key":"ref121","volume-title":"Dynamic Programming","author":"Bellman","year":"1957"},{"key":"ref122","doi-asserted-by":"publisher","DOI":"10.1109\/TIT.1967.1054010"},{"key":"ref123","doi-asserted-by":"publisher","DOI":"10.1109\/TASSP.1984.1164320"},{"key":"ref124","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2022-829"},{"key":"ref125","article-title":"Why does CTC result in Peaky behavior?","author":"Zeyer","year":"2021"},{"key":"ref126","doi-asserted-by":"publisher","DOI":"10.21437\/interspeech.2022-10854"},{"key":"ref127","doi-asserted-by":"publisher","DOI":"10.1109\/msp.2008.926652"},{"key":"ref128","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2021-1255"},{"key":"ref129","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP43922.2022.9746814"},{"key":"ref130","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP49357.2023.10094609"},{"key":"ref131","doi-asserted-by":"publisher","DOI":"10.1016\/S0167-6393(97)00029-0"},{"key":"ref132","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2001.940763"},{"key":"ref133","doi-asserted-by":"publisher","DOI":"10.1016\/S0167-6393(00)00035-2"},{"key":"ref134","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2009.4960445"},{"key":"ref135","doi-asserted-by":"publisher","DOI":"10.1109\/MSP.2012.2197232"},{"key":"ref136","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2019-2254"},{"key":"ref137","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2018.8461809"},{"key":"ref138","doi-asserted-by":"publisher","DOI":"10.21437\/interspeech.2020-1221"},{"key":"ref139","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2019.8682782"},{"key":"ref140","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2018.8461705"},{"key":"ref141","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2018.8462245"},{"key":"ref142","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2020-2675"},{"key":"ref143","doi-asserted-by":"publisher","DOI":"10.1162\/neco.2006.18.7.1527"},{"key":"ref144","doi-asserted-by":"publisher","DOI":"10.7551\/mitpress\/7503.003.0024"},{"key":"ref145","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2017.7952599"},{"key":"ref146","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P18-4022"},{"key":"ref147","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2018-1616"},{"key":"ref148","article-title":"A comprehensive analysis on attention models","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Zeyer","year":"2018"},{"key":"ref149","doi-asserted-by":"publisher","DOI":"10.21437\/interspeech.2016-82"},{"key":"ref150","doi-asserted-by":"publisher","DOI":"10.1109\/taslp.2019.2922832"},{"key":"ref151","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2008-672"},{"key":"ref152","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2013.6639090"},{"key":"ref153","article-title":"Multilingual end-to-end speech recognition with a single transformer on low-resource languages","author":"Zhou","year":"2018"},{"key":"ref154","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/N19-1009"},{"key":"ref155","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2020-2164"},{"key":"ref156","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2020-2831"},{"key":"ref157","doi-asserted-by":"publisher","DOI":"10.1109\/ASRU51503.2021.9687871"},{"key":"ref158","doi-asserted-by":"publisher","DOI":"10.1109\/JSTSP.2022.3182537"},{"key":"ref159","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2022-10937"},{"key":"ref160","article-title":"Introducing whisper - robust speech recognition via large-scale weak supervision","author":"Radford","year":"2022"},{"key":"ref161","doi-asserted-by":"publisher","DOI":"10.1007\/BF00332914"},{"key":"ref162","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2015.7178917"},{"key":"ref163","article-title":"Connectionist speech recognition: Status and prospects","volume-title":"Int. Conf. Sci. Inst.","author":"Renals","year":"1991"},{"key":"ref164","article-title":"QuickNet","volume-title":"Int. Conf. Swarm Intell.","author":"Johnson","year":"2004"},{"key":"ref165","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2013.6638963"},{"key":"ref166","article-title":"Decoupled weight decay regularization","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Loshchilov","year":"2019"},{"key":"ref167","article-title":"Dont decay the learning rate, increase the batch size","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Smith","year":"2018"},{"key":"ref168","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P18-1031"},{"key":"ref169","article-title":"Population based training of neural networks","author":"Jaderberg","year":"2017"},{"key":"ref170","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2021.3079209"},{"key":"ref171","doi-asserted-by":"publisher","DOI":"10.1016\/0010-0277(93)90058-4"},{"key":"ref172","doi-asserted-by":"publisher","DOI":"10.1145\/1553374.1553380"},{"key":"ref173","first-page":"173","article-title":"Deep speech 2: End-to-end speech recognition in english and mandarin","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Amodei","year":"2016"},{"key":"ref174","doi-asserted-by":"publisher","DOI":"10.21437\/interspeech.2020-1488"},{"key":"ref175","doi-asserted-by":"publisher","DOI":"10.1109\/TASLP.2020.2988423"},{"key":"ref176","doi-asserted-by":"publisher","DOI":"10.1016\/0041-5553(64)90137-5"},{"key":"ref177","first-page":"372","article-title":"A method of solving a convex programming problem with convergence rate O($\\frac{1}{k^{2}}$","volume":"27","author":"Nesterov","year":"1983","journal-title":"Sov. Math. Doklady"},{"key":"ref178","first-page":"1139","article-title":"On the importance of initialization and momentum in deep learning","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Sutskever","year":"2013"},{"key":"ref179","article-title":"Adam: A method for stochastic optimization","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Kingma","year":"2015"},{"key":"ref180","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2021-211"},{"key":"ref181","doi-asserted-by":"publisher","DOI":"10.1088\/1742-5468\/ac3a74"},{"key":"ref182","first-page":"950","article-title":"A simple weight decay can improve generalization","volume-title":"Proc. Neural Inf. Process. Syst.","author":"Krogh","year":"1991"},{"key":"ref183","doi-asserted-by":"publisher","DOI":"10.1109\/72.317730"},{"key":"ref184","doi-asserted-by":"publisher","DOI":"10.5555\/2986459.2986721"},{"key":"ref185","article-title":"Adding gradient noise improves learning for very deep networks","author":"Neelakantan","year":"2015"},{"key":"ref186","article-title":"Improving neural networks by preventing co-adaptation of feature detectors","author":"Hinton","year":"2012"},{"key":"ref187","doi-asserted-by":"publisher","DOI":"10.1145\/3065386"},{"key":"ref188","doi-asserted-by":"publisher","DOI":"10.5555\/3045390.3045502"},{"key":"ref189","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-46493-0_39"},{"key":"ref190","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2019-2702"},{"key":"ref191","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP39728.2021.9414594"},{"key":"ref192","first-page":"1058","article-title":"Regularization of neural networks using DropConnect","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Wan","year":"2013"},{"key":"ref193","article-title":"Zoneout: Regularizing RNNs by randomly preserving hidden activations","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Krueger","year":"2017"},{"key":"ref194","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.308"},{"key":"ref195","article-title":"Scheduled sampling for sequence prediction with recurrent neural networks","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Bengio","year":"2015"},{"key":"ref196","first-page":"4965","article-title":"Learning longer-term dependencies in RNNs with auxiliary losses","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Trinh","year":"2018"},{"key":"ref197","doi-asserted-by":"publisher","DOI":"10.1162\/neco.1990.2.4.490"},{"key":"ref198","article-title":"An analysis of neural language modeling at multiple scales","author":"Merity","year":"2018"},{"key":"ref199","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP39728.2021.9414483"},{"key":"ref200","doi-asserted-by":"publisher","DOI":"10.5555\/3045118.3045167"},{"key":"ref201","doi-asserted-by":"publisher","DOI":"10.1109\/ASRU.2013.6707748"},{"key":"ref202","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2015-711"},{"key":"ref203","article-title":"Vocal tract length perturbation (VTLP) improves speech recognition","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Jaitly","year":"2013"},{"key":"ref204","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2019.8683706"},{"key":"ref205","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP40776.2020.9053205"},{"key":"ref206","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2020-1778"},{"key":"ref207","doi-asserted-by":"publisher","DOI":"10.1109\/SLT.2018.8639619"},{"key":"ref208","doi-asserted-by":"publisher","DOI":"10.1109\/ASRU51503.2021.9688255"},{"key":"ref209","doi-asserted-by":"publisher","DOI":"10.1109\/icassp.2018.8462380"},{"key":"ref210","doi-asserted-by":"publisher","DOI":"10.21437\/ICSLP.1994-355"},{"key":"ref211","doi-asserted-by":"publisher","DOI":"10.1109\/TASL.2012.2226158"},{"key":"ref212","doi-asserted-by":"publisher","DOI":"10.1109\/icassp.2014.6854675"},{"key":"ref213","doi-asserted-by":"publisher","DOI":"10.21437\/interspeech.2016-391"},{"key":"ref214","doi-asserted-by":"publisher","DOI":"10.1109\/TASLP.2018.2848701"},{"key":"ref215","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2005.1415086"},{"key":"ref216","doi-asserted-by":"publisher","DOI":"10.1109\/slt.2018.8639684"},{"key":"ref217","doi-asserted-by":"publisher","DOI":"10.21437\/interspeech.2018-79"},{"key":"ref218","doi-asserted-by":"publisher","DOI":"10.21437\/Eurospeech.1993-494"},{"key":"ref219","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2010-10"},{"key":"ref220","article-title":"Towards consistent hybrid HMM acoustic modeling","author":"Raissi","year":"2021"},{"key":"ref221","article-title":"A systematic comparison of grapheme-based vs. phoneme-based label units for encoder-decoder-attention models","author":"Zeineldeen","year":"2020"},{"key":"ref222","doi-asserted-by":"publisher","DOI":"10.21437\/interspeech.2019-1780"},{"key":"ref223","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2020-1470"},{"key":"ref224","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2019-2680"},{"key":"ref225","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP40776.2020.9053573"},{"key":"ref226","doi-asserted-by":"publisher","DOI":"10.1109\/ASRU.2015.7404803"},{"key":"ref227","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/N19-1009"},{"key":"ref228","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2019-2858"},{"key":"ref229","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-642-24797-2_7"},{"key":"ref230","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2020-2404"},{"key":"ref231","first-page":"1403","article-title":"Imputer: Sequence modelling via imputation and dynamic programming","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Chan","year":"2020"},{"key":"ref232","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2020-1619"},{"key":"ref233","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP40776.2020.9054250"},{"key":"ref234","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2021-911"},{"key":"ref235","doi-asserted-by":"publisher","DOI":"10.1109\/ASRU51503.2021.9688157"},{"key":"ref236","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2020-1958"},{"key":"ref237","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/W17-3204"},{"key":"ref238","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P16-1008"},{"key":"ref239","doi-asserted-by":"publisher","DOI":"10.1109\/SLT.2018.8639693"},{"key":"ref240","article-title":"Label-synchronous neural transducer for end-to-end ASR","author":"Deng","year":"2023"},{"key":"ref241","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-02562-4"},{"key":"ref242","doi-asserted-by":"publisher","DOI":"10.1109\/89.279287"},{"key":"ref243","doi-asserted-by":"publisher","DOI":"10.1109\/5.880081"},{"key":"ref244","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2014.6854829"},{"key":"ref245","article-title":"LSTM language models for LVCSR in first-pass decoding and lattice-rescoring","author":"Beck","year":"2019"},{"key":"ref246","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP40776.2020.9053040"},{"key":"ref247","article-title":"First-pass large vocabulary continuous speech recognition using bi-directional recurrent DNNs","author":"Hannun","year":"2014"},{"key":"ref248","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2019.8683510"},{"key":"ref249","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP40776.2020.9054476"},{"key":"ref250","article-title":"RNN-T for latency controlled ASR with improved beam search","author":"Jain","year":"2019"},{"key":"ref251","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2020-2638"},{"key":"ref252","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2021-1556"},{"key":"ref253","doi-asserted-by":"publisher","DOI":"10.21437\/interspeech.2019-2018"},{"key":"ref254","doi-asserted-by":"publisher","DOI":"10.1109\/SLT48900.2021.9383517"},{"key":"ref255","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2017.7953252"},{"key":"ref256","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2017-1296"},{"key":"ref257","doi-asserted-by":"publisher","DOI":"10.1109\/icassp.2018.8462682"},{"key":"ref258","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP39728.2021.9414716"},{"key":"ref259","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2019-2860"},{"key":"ref260","doi-asserted-by":"publisher","DOI":"10.1109\/ASRU.2017.8268948"},{"key":"ref261","doi-asserted-by":"publisher","DOI":"10.1109\/ASRU46091.2019.9003968"},{"key":"ref262","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2019-3018"},{"key":"ref263","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2019.8683531"},{"key":"ref264","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2019-1341"},{"key":"ref265","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2021-1983"},{"key":"ref266","article-title":"U2++: Unified two-pass bidirectional end-to-end model for speech recognition","author":"Wu","year":"2021"},{"key":"ref267","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2019-2667"},{"key":"ref268","doi-asserted-by":"publisher","DOI":"10.1109\/LSP.2020.3036335"},{"key":"ref269","first-page":"265","article-title":"TensorFlow: A system for large-scale machine learning","volume-title":"Proc. 12th USENIX Symp. Operating Syst. Des. Implementation","author":"Abadi","year":"2016"},{"key":"ref270","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/N19-4009"},{"key":"ref271","article-title":"Lingvo: A modular and scalable framework for sequence-to-sequence modeling","author":"Shen","year":"2019"},{"key":"ref272","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2017.7953177"},{"key":"ref273","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2019-2460"},{"key":"ref274","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2019.8682500"},{"key":"ref275","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP40776.2020.9054545"},{"key":"ref276","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2014.6854827"},{"key":"ref277","doi-asserted-by":"publisher","DOI":"10.21437\/interspeech.2019-2798"},{"key":"ref278","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP40776.2020.9053010"},{"key":"ref279","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D16-1158"},{"key":"ref280","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/W18-6322"},{"key":"ref281","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D19-1331"},{"key":"ref282","doi-asserted-by":"publisher","DOI":"10.1109\/79.790985"},{"key":"ref283","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.1999.759741"},{"key":"ref284","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP40776.2020.9054249"},{"key":"ref285","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2019-2719"},{"key":"ref286","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2019-3135"},{"key":"ref287","first-page":"3163","article-title":"Evaluation of feature-space speaker adaptation for end-to-end acoustic models","volume-title":"Proc. 11th Int. Conf. Lang. Resour. Eval.","author":"Tomashenko","year":"2018"},{"key":"ref288","doi-asserted-by":"publisher","DOI":"10.3115\/981863.981904"},{"key":"ref289","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2010-343"},{"key":"ref290","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2012-65"},{"key":"ref291","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D16-1123"},{"key":"ref292","first-page":"933","article-title":"Language modeling with gated convolutional networks","volume-title":"Proc. 34th Int. Conf. Mach. Learn.","author":"Dauphin","year":"2017"},{"key":"ref293","article-title":"Fully convolutional speech recognition","author":"Zeghidour","year":"2018"},{"key":"ref294","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2019-3107"},{"key":"ref295","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v33i01.33013159"},{"key":"ref296","doi-asserted-by":"publisher","DOI":"10.21437\/interspeech.2019-2225"},{"key":"ref297","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P19-1285"},{"key":"ref298","doi-asserted-by":"publisher","DOI":"10.1109\/5.58337"},{"key":"ref299","first-page":"6154","article-title":"Compressive transformers for long-range sequence modelling","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Rae","year":"2020,"},{"key":"ref300","article-title":"On using monolingual corpora in neural machine translation","author":"Gulcehre","year":"2015"},{"key":"ref301","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2018-1392"},{"key":"ref302","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2019.8682490"},{"key":"ref303","doi-asserted-by":"publisher","DOI":"10.1109\/ASRU46091.2019.9003790"},{"key":"ref304","doi-asserted-by":"publisher","DOI":"10.1109\/SLT48900.2021.9383515"},{"key":"ref305","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP43922.2022.9746948"},{"key":"ref306","first-page":"4171","article-title":"BERT: Pre-training of deep bidirectional transformers for language understanding","volume-title":"Proc. Assoc. Comput. Linguistics","author":"Devlin","year":"2019"},{"key":"ref307","article-title":"Language models are unsupervised multitask learners","author":"Radford","year":"2019"},{"key":"ref308","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2020.acl-main.240"},{"key":"ref309","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P19-1107"},{"key":"ref310","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2021-1510"},{"key":"ref311","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.1983.4767370"},{"key":"ref312","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.92.22.9956"},{"key":"ref313","doi-asserted-by":"publisher","DOI":"10.1016\/S0167-6393(01)00041-3"},{"key":"ref314","doi-asserted-by":"publisher","DOI":"10.1109\/TASLP.2015.2400218"},{"key":"ref315","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2016.7472827"},{"key":"ref316","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.1992.225858"},{"key":"ref317","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2015.7178964"},{"key":"ref318","doi-asserted-by":"publisher","DOI":"10.1109\/ASRU46091.2019.9004025"},{"key":"ref319","doi-asserted-by":"publisher","DOI":"10.1109\/TASLP.2021.3122291"},{"key":"ref320","article-title":"End-to-end ASR: From supervised to semi-supervised learning with modern architectures","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Synnaeve","year":"2020"},{"key":"ref321","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2021-337"},{"key":"ref322","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP40776.2020.9052942"},{"key":"ref323","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP40776.2020.9054345"},{"key":"ref324","doi-asserted-by":"publisher","DOI":"10.1109\/SLT54892.2023.10022656"},{"key":"ref325","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2019-2162"},{"key":"ref326","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2018.8462105"},{"key":"ref327","doi-asserted-by":"publisher","DOI":"10.1109\/ASRU46091.2019.9004027"},{"key":"ref328","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2020-3016"},{"key":"ref329","doi-asserted-by":"publisher","DOI":"10.1109\/LSP.2020.3031480"},{"key":"ref330","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP39728.2021.9414560"},{"key":"ref331","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP39728.2021.9413535"},{"key":"ref332","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP40776.2020.9054188"},{"key":"ref333","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP39728.2021.9413899"},{"key":"ref334","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2021-206"},{"key":"ref335","article-title":"SLAM: A unified encoder for speech and language modeling via speech-text joint pre-training","author":"Bapna","year":"2021"},{"key":"ref336","article-title":"mSLAM: Massively Multilingual Joint Pre-Training for Speech and text","author":"Bapna","year":"2022"},{"key":"ref337","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2022.acl-long.105"},{"key":"ref338","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2021.naacl-main.152"},{"key":"ref339","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2022.acl-long.393"},{"key":"ref340","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP43922.2022.9747555"},{"key":"ref341","doi-asserted-by":"publisher","DOI":"10.1109\/SLT54892.2023.10022774"},{"key":"ref342","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2019.8683307"},{"key":"ref343","first-page":"2632","article-title":"Multichannel end-to-end speech recognition","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Ochiai","year":"2017"},{"key":"ref344","doi-asserted-by":"publisher","DOI":"10.1561\/116.00000050"}],"container-title":["IEEE\/ACM Transactions on Audio, Speech, and Language Processing"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx7\/6570655\/10304349\/10301513.pdf?arnumber=10301513","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,4,12]],"date-time":"2024-04-12T02:55:17Z","timestamp":1712890517000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/10301513\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024]]},"references-count":344,"URL":"https:\/\/doi.org\/10.1109\/taslp.2023.3328283","relation":{},"ISSN":["2329-9290","2329-9304"],"issn-type":[{"value":"2329-9290","type":"print"},{"value":"2329-9304","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024]]}}}