VietASR: Achieving Industry-level Vietnamese ASR with 50-hour labeled data and Large-Scale Speech Pretraining
Not on the map. Why?
Abstract (arXiv)
Automatic speech recognition (ASR) has made remarkable progress but heavily relies on large-scale labeled data, which is scarce for low-resource languages like Vietnamese. While existing systems such as Whisper, USM, and MMS achieve promising performance, their efficacy remains inadequate in terms of training costs, latency, and accessibility. To address these issues, we propose VietASR, a novel ASR training pipeline that leverages vast amounts of unlabeled data and a small set of labeled data. Through multi-iteration ASR-biased self-supervised learning on a large-scale unlabeled dataset, VietASR offers a cost-effective and practical solution for enhancing ASR performance. Experiments demonstrate that pre-training on 70,000-hour unlabeled data and fine-tuning on merely 50-hour labeled data yield a lightweight but powerful ASR model. It outperforms Whisper Large-v3 and commercial ASR systems on real-world data. Our code and models will be open-sourced to facilitate research in low-resource ASR.
Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Builds on 28 of its 33 references that are in the map. First posted on arXiv 23 May 2025.
Builds on
28 of 33 references.
References · 28
- 2017Attention Is All You Needcited by 60,361
- 2022Robust Speech Recognition via Large-Scale Weak Supervisioncited by 4,057
- 2020wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representationscited by 3,905
- 2015Neural Machine Translation of Rare Words with Subword Unitscited by 2,918
- 2021HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Unitscited by 2,342
- 2020Conformer: Convolution-augmented Transformer for Speech Recognitioncited by 1,863
- 2019Common Voice: A Massively-Multilingual Speech Corpuscited by 1,384
- 2013Speech Recognition with Deep Recurrent Neural Networkscited by 669
- 2021XLS-R: Self-supervised Cross-lingual Speech Representation Learning at Scalecited by 601
- 2022FLEURS: Few-shot Learning Evaluation of Universal Representations of Speechcited by 469
- 2023Scaling Speech Technology to 1,000+ Languagescited by 411
- 2023Google USM: Scaling Automatic Speech Recognition Beyond 100 Languagescited by 231
16 more
- 2021Recent Advances in End-to-End Automatic Speech Recognitioncited by 140
- 2023FunASR: A Fundamental End-to-End Speech Recognition Toolkitcited by 95
- 2023Zipformer: A faster and better encoder for automatic speech recognitioncited by 80
- 2022Pruned RNN-T for fast, memory-efficient ASR trainingcited by 46
- 2024GigaSpeech 2: An Evolving, Large-Scale and Multi-domain ASR Corpus for Low-Resource Languages with Automated Crawling, Transcription and Refinementcited by 34
- 2023Reducing Barriers to Self-Supervised Learning: HuBERT Pre-training with Academic Computecited by 21
- 2022MelHuBERT: A simplified HuBERT on Mel spectrogramscited by 16
- 2022Speech Pre-training with Acoustic Piececited by 14
- 2022Supervision-Guided Codebooks for Masked Prediction in Speech Pre-trainingcited by 14
- 2024Anatomy of Industrial Scale Multilingual ASRcited by 14
- 2022Fast and parallel decoding for transducercited by 12
- 2023Fast-HuBERT: An Efficient Training Framework for Self-Supervised Speech Representation Learningcited by 9
- 2023Pushing the Limits of Unsupervised Unit Discovery for SSL Speech Representationcited by 7
- 2024k2SSL: A Faster and Better Framework for Self-Supervised Speech Representation Learningcited by 6
- 2022CTCBERT: Advancing Hidden-unit BERT with CTC Objectivescited by 5
- 2022Biased Self-supervised learning for ASRcited by 3
Built on by
Citing papers per year: 2026: 3.
3 papers.