{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,11]],"date-time":"2026-07-11T16:26:07Z","timestamp":1783787167482,"version":"3.55.0"},"reference-count":124,"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-nc-nd\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Access"],"published-print":{"date-parts":[[2024]]},"DOI":"10.1109\/access.2024.3428364","type":"journal-article","created":{"date-parts":[[2024,7,15]],"date-time":"2024-07-15T18:23:17Z","timestamp":1721067797000},"page":"98835-98855","source":"Crossref","is-referenced-by-count":3,"title":["Unveiling the Linguistic Capabilities of a Self-Supervised Speech Model Through Cross-Lingual Benchmark and Layer- Wise Similarity Analysis"],"prefix":"10.1109","volume":"12","author":[{"ORCID":"https:\/\/orcid.org\/0009-0003-4322-4127","authenticated-orcid":false,"given":"Takanori","family":"Ashihara","sequence":"first","affiliation":[{"name":"NTT Corporation, Yokosuka-shi, Kanagawa-ken, Japan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5175-7834","authenticated-orcid":false,"given":"Marc","family":"Delcroix","sequence":"additional","affiliation":[{"name":"NTT Corporation, Yokosuka-shi, Kanagawa-ken, Japan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yusuke","family":"Ijima","sequence":"additional","affiliation":[{"name":"NTT Corporation, Yokosuka-shi, Kanagawa-ken, Japan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4354-6690","authenticated-orcid":false,"given":"Makio","family":"Kashino","sequence":"additional","affiliation":[{"name":"NTT Corporation, Yokosuka-shi, Kanagawa-ken, Japan"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1561\/2200000006"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1038\/nature14539"},{"key":"ref3","volume-title":"Deep Learning","author":"Goodfellow","year":"2016"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1109\/MSP.2012.2205597"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1080\/02763869.2018.1404391"},{"key":"ref6","first-page":"28492","article-title":"Robust speech recognition via large-scale weak supervision","volume-title":"Proc. Int. Conf. Mach. Learn.","volume":"22","author":"Radford"},{"key":"ref7","article-title":"GPT-4 technical report","volume-title":"arXiv:2303.08774","author":"Achiam","year":"2023"},{"key":"ref8","volume-title":"Video Generation Models as World Simulators","author":"Brooks","year":"2024"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2013.50"},{"key":"ref10","article-title":"On the opportunities and risks of foundation models","author":"Bommasani","year":"2021","journal-title":"arXiv:2108.07258"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1109\/MSP.2021.3134634"},{"key":"ref12","article-title":"A cookbook of self-supervised learning","author":"Balestriero","year":"2023","journal-title":"arXiv:2304.12210"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1162\/tacl_a_00349"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1007\/s11431-020-1647-3"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2020.emnlp-main.608"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1145\/3605943"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2020.2992393"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00202"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1109\/JSTSP.2022.3207050"},{"key":"ref20","article-title":"A brief overview of unsupervised neural speech representation learning","volume-title":"Proc. 2nd Workshop Self-Supervised Learn. Audio Speech Process.","author":"Borgholt"},{"key":"ref21","article-title":"Deep representation learning in speech processing: Challenges, recent advances, and future trends","author":"Latif","year":"2020","journal-title":"arXiv:2001.00378"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1016\/j.patter.2022.100616"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2022.3170559"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1038\/s41551-022-00914-1"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1109\/TKDE.2021.3090866"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1810.04805"},{"key":"ref27","article-title":"Neural discrete representation learning","volume-title":"Proc. Adv. Neural Inf. Process. Syst. (NeurIPS)","volume":"30","author":"van den Oord"},{"key":"ref28","article-title":"vq-wav2vec: Self-supervised learning of discrete speech representations","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR)","author":"Baevski"},{"key":"ref29","first-page":"12449","article-title":"Wav2vec 2.0: A framework for self-supervised learning of speech representations","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst. (NIPS)","volume":"33","author":"Baevski"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1109\/TASLP.2021.3122291"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1109\/JSTSP.2022.3188113"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1109\/ASRU51503.2021.9688137"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2021-1280"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP43922.2022.9747814"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2022-10019"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2023.3263585"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP43922.2022.9746303"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2023-1578"},{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP39728.2021.9415079"},{"key":"ref40","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSPW59220.2023.10193459"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP43922.2022.9747674"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP43922.2022.9746137"},{"key":"ref43","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.acl-long.496"},{"key":"ref44","doi-asserted-by":"publisher","DOI":"10.1109\/SLT54892.2023.10022399"},{"key":"ref45","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/W18-5446"},{"key":"ref46","doi-asserted-by":"publisher","DOI":"10.21437\/interspeech.2021-1775"},{"key":"ref47","article-title":"SLAM: A unified encoder for speech and language modeling via speech-text joint pre-training","author":"Bapna","year":"2021","journal-title":"arXiv:2110.10329"},{"key":"ref48","article-title":"MSLAM: Massively multilingual joint pre-training for speech and text","author":"Bapna","year":"2022","journal-title":"arXiv:2202.01374"},{"key":"ref49","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP49357.2023.10096923"},{"key":"ref50","doi-asserted-by":"publisher","DOI":"10.1109\/ASRU57964.2023.10389660"},{"key":"ref51","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2021-329"},{"key":"ref52","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2022-143"},{"key":"ref53","article-title":"Google USM: Scaling automatic speech recognition beyond 100 languages","author":"Zhang","year":"2023","journal-title":"arXiv:2303.01037"},{"issue":"97","key":"ref54","first-page":"1","article-title":"Scaling speech technology to 1,000+ languages","volume":"25","author":"Pratap","year":"2024","journal-title":"J. Mach. Learn. Res."},{"key":"ref55","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2023-1823"},{"key":"ref56","doi-asserted-by":"publisher","DOI":"10.5715\/jnlp.29.711"},{"key":"ref57","doi-asserted-by":"publisher","DOI":"10.1109\/ASRU51503.2021.9688093"},{"key":"ref58","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP49357.2023.10096149"},{"key":"ref59","article-title":"What do self-supervised speech models know about words?","author":"Pasad","year":"2023","journal-title":"arXiv:2307.00162"},{"key":"ref60","article-title":"Representation learning with contrastive predictive coding","author":"van den Oord","year":"2018","journal-title":"arXiv:1807.03748"},{"key":"ref61","first-page":"1298","article-title":"data2vec: A general framework for self-supervised learning in speech, vision and language","volume-title":"Proc. Int. Conf. Mach. Learn. (ICML)","volume":"162","author":"Baevski"},{"key":"ref62","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2022.acl-long.427"},{"key":"ref63","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2022.acl-long.580"},{"key":"ref64","doi-asserted-by":"publisher","DOI":"10.1109\/ASRU57964.2023.10389699"},{"key":"ref65","article-title":"EMO-SUPERB: An in-depth look at speech emotion recognition","author":"Wu","year":"2024","journal-title":"arXiv:2402.13018"},{"key":"ref66","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2023-1316"},{"key":"ref67","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP48485.2024.10445941"},{"key":"ref68","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2020-1242"},{"key":"ref69","article-title":"Task agnostic and task specific self-supervised learning from speech with LeBenchmark","volume-title":"Proc. Neural Inf. Process. Syst. Track Datasets Benchmarks","volume":"1","author":"Evain"},{"key":"ref70","doi-asserted-by":"publisher","DOI":"10.1016\/j.csl.2024.101622"},{"key":"ref71","first-page":"125","article-title":"HEAR: Holistic evaluation of audio representations","volume-title":"Proc. NeurIPS Competitions Demonstrations Track","volume":"176","author":"Turian"},{"key":"ref72","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2022-10007"},{"key":"ref73","article-title":"ASR-GLUE: A new multi-task benchmark for ASR-robust natural language understanding","author":"Feng","year":"2021","journal-title":"arXiv:2108.13048"},{"key":"ref74","article-title":"The zero resource speech benchmark 2021: Metrics and baselines for unsupervised spoken language modeling","author":"Nguyen","year":"2020","journal-title":"arXiv:2011.11588"},{"key":"ref75","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2021-1755"},{"key":"ref76","doi-asserted-by":"publisher","DOI":"10.1109\/JSTSP.2022.3200909"},{"key":"ref77","first-page":"63483","article-title":"Textually pretrained speech language models","volume-title":"Proc. Adv. Neural Inf. Process. Syst. (NeurIPS)","volume":"36","author":"Hassid"},{"key":"ref78","first-page":"1336","article-title":"On generative spoken language modeling from raw audio","volume":"9","author":"Lakhotia","year":"2021","journal-title":"Trans. Assoc. Comput. Linguistics"},{"key":"ref79","article-title":"SuperGLUE: A stickier benchmark for general-purpose language understanding systems","volume-title":"Proc. Adv. Neural Inf. Process. Syst. (NeurIPS)","volume":"32","author":"Wang"},{"key":"ref80","article-title":"Beyond the imitation game: Quantifying and extrapolating the capabilities of language models","year":"2023","journal-title":"Trans. Mach. Learn. Res."},{"key":"ref81","doi-asserted-by":"publisher","DOI":"10.1162\/tacl_a_00290"},{"key":"ref82","first-page":"1631","article-title":"Recursive deep models for semantic compositionality over a sentiment treebank","volume-title":"Proc. Conf. Empir. Methods Nat. Lang. Process.","author":"Socher"},{"key":"ref83","article-title":"Automatically constructing a corpus of sentential paraphrases","volume-title":"Proc. Int. Workshop Paraphrasing (IWP)","author":"Dolan"},{"key":"ref84","volume-title":"First Quora Dataset Release: Question Pairs","author":"Iyer","year":"2017"},{"key":"ref85","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/s17-2001"},{"key":"ref86","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/N18-1101"},{"key":"ref87","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D16-1264"},{"key":"ref88","doi-asserted-by":"publisher","DOI":"10.1007\/11736790_9"},{"key":"ref89","article-title":"The second PASCAL recognising textual entailment challenge","volume-title":"Proc. 2nd PASCAL Challenges Workshop Recognizing Textual Entailment","author":"Bar-Haim"},{"key":"ref90","doi-asserted-by":"publisher","DOI":"10.3115\/1654536.1654538"},{"key":"ref91","first-page":"8","article-title":"The fifth PASCAL recognizing textual entailment challenge","volume-title":"Proc. TAC","volume":"7","author":"Bentivogli"},{"key":"ref92","article-title":"The Winograd schema challenge","volume-title":"Proc. Int. Conf. Principles Knowl. Represent. Reasoning","author":"Levesque"},{"key":"ref93","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2020.coling-main.419"},{"key":"ref94","first-page":"2479","article-title":"FlauBERT: Unsupervised language model pre-training for French","volume-title":"Proc. 12th Lang. Resour. Eval. Conf.","author":"Le"},{"key":"ref95","article-title":"KLUE: Korean language understanding evaluation","volume-title":"Proc. Neural Inf. Process. Syst. Track Datasets Benchmarks","volume":"1","author":"Park"},{"key":"ref96","article-title":"A survey on neural speech synthesis","author":"Tan","year":"2021","journal-title":"arXiv:2106.15561"},{"key":"ref97","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.1907.11692"},{"key":"ref98","article-title":"XLNet: Generalized autoregressive pretraining for language understanding","volume-title":"Proc. Adv. Neural Inf. Process. Syst. (NeurIPS)","volume":"32","author":"Yang"},{"key":"ref99","first-page":"3519","article-title":"Similarity of neural network representations revisited","volume-title":"Proc. Int. Conf. Mach. Learn.","volume":"97","author":"Kornblith"},{"key":"ref100","article-title":"Do wide and deep networks learn the same things? Uncovering how neural network representations vary with width and depth","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR)","author":"Nguyen"},{"key":"ref101","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2020.acl-main.422"},{"key":"ref102","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP39728.2021.9414321"},{"key":"ref103","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2021-1516"},{"key":"ref104","article-title":"SKILL: Similarity-aware knowledge distillation for speech self-supervised learning","author":"Zampierin","year":"2024","journal-title":"arXiv:2402.16830"},{"key":"ref105","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP48485.2024.10446422"},{"issue":"1","key":"ref106","first-page":"1393","article-title":"Feature selection via dependence maximization","volume":"13","author":"Song","year":"2012","journal-title":"J. Mach. Learn. Res."},{"key":"ref107","article-title":"HuggingFaces transformers: State-of-the-art natural language processing","author":"Wolf","year":"2019","journal-title":"arXiv:1910.03771"},{"key":"ref108","first-page":"5530","article-title":"Conditional variational autoencoder with adversarial learning for end-to-end text-to-speech","volume-title":"Proc. Int. Conf. Mach. Learn.","volume":"139","author":"Kim"},{"key":"ref109","article-title":"JSUT corpus: Free large-scale Japanese speech corpus for end-to-end speech synthesis","author":"Sonobe","year":"2017","journal-title":"arXiv:1711.00354"},{"key":"ref110","doi-asserted-by":"publisher","DOI":"10.21437\/interspeech.2018-1456"},{"key":"ref111","article-title":"Corpus of spontaneous Japanese: Its design and evaluation","volume-title":"Proc. ISCA\/IEEE Workshop Spontaneous Speech Process. Recognit.","author":"Maekawa"},{"key":"ref112","first-page":"1134","article-title":"ReazonSpeech: A free and massive corpus for Japanese ASR","volume-title":"Proc. Annu. Meetings Assoc. Natural Lang. Process.","author":"Yin"},{"key":"ref113","first-page":"1418","article-title":"Frustratingly easy performance improvements for low-resource setups: A tale on BERT and segment embeddings","volume-title":"Proc. Lang. Resour. Eval. Conf. (LREC)","author":"van der Goot"},{"key":"ref114","first-page":"23102","article-title":"COCO-LM: Correcting and contrasting text sequences for language model pretraining","volume-title":"Proc. Adv. Neural Inf. Process. Syst. (NeurIPS)","volume":"34","author":"Meng"},{"key":"ref115","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP49357.2023.10096318"},{"key":"ref116","doi-asserted-by":"crossref","DOI":"10.6028\/NIST.IR.4930","volume-title":"DARPA TIMIT acoustic-phonetic continuous speech corpus CD-ROM. NIST speech disc 1\u20131.1","author":"Garofolo","year":"1993"},{"key":"ref117","doi-asserted-by":"publisher","DOI":"10.1250\/ast.20.199"},{"key":"ref118","doi-asserted-by":"publisher","DOI":"10.1109\/ICSDA.2017.8384449"},{"key":"ref119","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2019-2821"},{"key":"ref120","doi-asserted-by":"publisher","DOI":"10.1109\/ASRU51503.2021.9688253"},{"key":"ref121","article-title":"Scaling laws for neural language models","author":"Kaplan","year":"2020","journal-title":"arXiv:2001.08361"},{"key":"ref122","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2021-1935"},{"key":"ref123","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2021-236"},{"key":"ref124","doi-asserted-by":"publisher","DOI":"10.1016\/j.csl.2019.101027"}],"container-title":["IEEE Access"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/6287639\/10380310\/10597571.pdf?arnumber=10597571","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,7,25]],"date-time":"2024-07-25T17:37:30Z","timestamp":1721929050000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/10597571\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024]]},"references-count":124,"URL":"https:\/\/doi.org\/10.1109\/access.2024.3428364","relation":{},"ISSN":["2169-3536"],"issn-type":[{"value":"2169-3536","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024]]}}}