Continued Pretraining for Low-Resource Swahili ASR: Achieving State-of-the-Art Performance with Minimal Labeled Data

• URL: http://arxiv.org/abs/2603.11378v1
• Date: Wed, 11 Mar 2026 23:45:06 GMT
• Title: Continued Pretraining for Low-Resource Swahili ASR: Achieving State-of-the-Art Performance with Minimal Labeled Data
• Abstract summary: We adapt wav2vec2-bert-2.0 to Swahili automatic speech recognition.<n>With 20,000 labeled samples, we achieve 3.24% WER on Common Voice Swahili.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: We investigate continued pretraining (CPT) for adapting wav2vec2-bert-2.0 to Swahili automatic speech recognition (ASR). Our approach combines unlabeled audio with limited labeled data through pseudo-labeled CPT followed by supervised finetuning. With 20,000 labeled samples, we achieve 3.24% WER on Common Voice Swahili-an 82% relative improvement over the baseline. This result surpasses the best previously reported academic system (8.3% WER from XLS-R) by 61% relative improvement. We provide concrete data requirements and a replicable methodology applicable to other low-resource languages.

Related papers

• Selective Augmentation: Improving Universal Automatic Phonetic Transcription via G2P Bootstrapping [0.0]
  We propose the bootstrapping approach Selective Augmentation to improve the available training transcriptions.<n>We exemplarily show that we could increase the accuracy of an existing feature (plosive voicing) and add a new feature (plosive aspiration)<n>We describe intrinsic challenges of the evaluation and develop objective metrics to determine the success.
  arXiv  Detail & Related papers  (2026-04-29T21:19:59Z)
• Doing More with Less: Data Augmentation for Sudanese Dialect Automatic Speech Recognition [0.0]
  This paper presents a study of data augmentation techniques for fine-tuning OpenAI Whisper models.<n>It establishes the first benchmark for the Sudanese dialect.
  arXiv  Detail & Related papers  (2026-01-11T08:28:31Z)
• BRAVEn: Improving Self-Supervised Pre-training for Visual and Auditory Speech Recognition [72.51848069125822]
  We propose BRAVEn, an extension to the RAVEn method, which learns speech representations entirely from raw audio-visual data. Our modifications to RAVEn enable BRAVEn to achieve state-of-the-art results among self-supervised methods. Our results suggest that readily available unlabelled audio-visual data can largely replace costly transcribed data.
  arXiv  Detail & Related papers  (2024-04-02T16:48:20Z)
• Multilingual Audio-Visual Speech Recognition with Hybrid CTC/RNN-T Fast Conformer [59.57249127943914]
  We present a multilingual Audio-Visual Speech Recognition model incorporating several enhancements to improve performance and audio noise robustness. We increase the amount of audio-visual training data for six distinct languages, generating automatic transcriptions of unlabelled multilingual datasets. Our proposed model achieves new state-of-the-art performance on the LRS3 dataset, reaching WER of 0.8%.
  arXiv  Detail & Related papers  (2024-03-14T01:16:32Z)
• Strategies for improving low resource speech to text translation relying on pre-trained ASR models [59.90106959717875]
  This paper presents techniques and findings for improving the performance of low-resource speech to text translation (ST) We conducted experiments on both simulated and real-low resource setups, on language pairs English - Portuguese, and Tamasheq - French respectively.
  arXiv  Detail & Related papers  (2023-05-31T21:58:07Z)
• Auto-AVSR: Audio-Visual Speech Recognition with Automatic Labels [100.43280310123784]
  We investigate the use of automatically-generated transcriptions of unlabelled datasets to increase the training set size. We demonstrate that increasing the size of the training set, a recent trend in the literature, leads to reduced WER despite using noisy transcriptions. The proposed model achieves new state-of-the-art performance on AV-ASR on LRS2 and LRS3.
  arXiv  Detail & Related papers  (2023-03-25T00:37:34Z)
• Self-Training with Purpose Preserving Augmentation Improves Few-shot Generative Dialogue State Tracking [14.709084509818474]
  In dialogue state tracking (DST), labeling the dataset involves considerable human labor. We propose a new self-training framework for few-shot generative DST that utilize unlabeled data.
  arXiv  Detail & Related papers  (2022-11-17T07:13:58Z)
• Improving Low-Resource Speech Recognition with Pretrained Speech Models: Continued Pretraining vs. Semi-Supervised Training [6.523198497365586]
  Self-supervised Transformer based models, such as wav2vec 2.0 and HuBERT, have produced significant improvements over existing approaches to automatic speech recognition (ASR) We show CoPT results in word error rates (WERs) equal to or slightly better than using semi-supervised training (SST) In addition, we show that using the CoPT model for pseudo-labeling, and using these labels in SST, results in further improvements in WER.
  arXiv  Detail & Related papers  (2022-07-01T21:02:51Z)
• Semi-supervised transfer learning for language expansion of end-to-end speech recognition models to low-resource languages [19.44975351652865]
  We propose a three-stage training methodology to improve the speech recognition accuracy of low-resource languages. We leverage a well-trained English model, unlabeled text corpus, and unlabeled audio corpus using transfer learning, TTS augmentation, and SSL respectively. Overall, our two-pass speech recognition system with a Monotonic Chunkwise Attention (MoA) in the first pass achieves a WER reduction of 42% relative to the baseline.
  arXiv  Detail & Related papers  (2021-11-19T05:09:16Z)
• Pushing the Limits of Semi-Supervised Learning for Automatic Speech Recognition [97.44056170380726]
  We employ a combination of recent developments in semi-supervised learning for automatic speech recognition to obtain state-of-the-art results on LibriSpeech. We carry out noisy student training with SpecAugment using giant Conformer models pre-trained using wav2vec 2.0 pre-training. We are able to achieve word-error-rates (WERs) 1.4%/2.6% on the LibriSpeech test/test-other sets against the current state-of-the-art WERs 1.7%/3.3%.
  arXiv  Detail & Related papers  (2020-10-20T17:58:13Z)
• Towards a Competitive End-to-End Speech Recognition for CHiME-6 Dinner Party Transcription [73.66530509749305]
  In this paper, we argue that, even in difficult cases, some end-to-end approaches show performance close to the hybrid baseline. We experimentally compare and analyze CTC-Attention versus RNN-Transducer approaches along with RNN versus Transformer architectures. Our best end-to-end model based on RNN-Transducer, together with improved beam search, reaches quality by only 3.8% WER abs. worse than the LF-MMI TDNN-F CHiME-6 Challenge baseline.
  arXiv  Detail & Related papers  (2020-04-22T19:08:33Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.