# Low-Resource Speech-to-Text Translation

arXiv: https://arxiv.org/abs/1803.09164 · HTML: https://mlsanity.com/paper/1803.09164/

## Key facts

- Cited by 12 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 24 Mar 2018.

## Built on by (most cited first, 12 of 12)

- [RealTranS: End-to-End Simultaneous Speech Translation with Convolutional Weighted-Shrinking Transformer](https://mlsanity.com/paper/2106.04833/) (2021)
- [How "Real" is Your Real-Time Simultaneous Speech-to-Text Translation System?](https://mlsanity.com/paper/2412.18495/) (2024)
- [Simple and Effective Unsupervised Speech Translation](https://mlsanity.com/paper/2210.10191/) (2022)
- [Efficient Monotonic Multihead Attention](https://mlsanity.com/paper/2312.04515/) (2023)
- [Discrete Cross-Modal Alignment Enables Zero-Shot Speech Translation](https://mlsanity.com/paper/2210.09556/) (2022)
- [Zero-shot Speech Translation](https://mlsanity.com/paper/2107.06010/) (2021)
- [Tackling data scarcity in speech translation using zero-shot multilingual machine translation techniques](https://mlsanity.com/paper/2201.11172/) (2022)
- [Transfer learning from High-Resource to Low-Resource Language Improves Speech Affect Recognition Classification Accuracy](https://mlsanity.com/paper/2103.11764/) (2021)
- [Where Are You From? Let Me Guess! Subdialect Recognition of Speeches in Sorani Kurdish](https://mlsanity.com/paper/2404.00124/) (2024)
- [Exploiting Vulnerabilities in Speech Translation Systems through Targeted Adversarial Attacks](https://mlsanity.com/paper/2503.00957/) (2025)
- [Selective Data Augmentation for Robust Speech Translation](https://mlsanity.com/paper/2304.03169/) (2023)
- [LIMBA: An Open-Source Framework for the Preservation and Valorization of Low-Resource Languages using Generative Models](https://mlsanity.com/paper/2411.13453/) (2024)

## Abstract (arXiv)

Speech-to-text translation has many potential applications for low-resource languages, but the typical approach of cascading speech recognition with machine translation is often impossible, since the transcripts needed to train a speech recognizer are usually not available for low-resource languages. Recent work has found that neural encoder-decoder models can learn to directly translate foreign speech in high-resource scenarios, without the need for intermediate transcription. We investigate whether this approach also works in settings where both data and computation are limited. To make the approach efficient, we make several architectural changes, including a change from character-level to word-level decoding. We find that this choice yields crucial speed improvements that allow us to train with fewer computational resources, yet still performs well on frequent words. We explore models trained on between 20 and 160 hours of data, and find that although models trained on less data have considerably lower BLEU scores, they can still predict words with relatively high precision and recall---around 50% for a model trained on 50 hours of data, versus around 60% for the full 160 hour model. Thus, they may still be useful for some low-resource scenarios.