# Phone Features Improve Speech Translation

arXiv: https://arxiv.org/abs/2005.13681 · HTML: https://mlsanity.com/paper/2005.13681/

## Key facts

- Cited by 11 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 27 May 2020.

## Built on by (most cited first, 11 of 11)

- [The Multilingual TEDx Corpus for Speech Recognition and Translation](https://mlsanity.com/paper/2102.01757/) (2021)
- [Improving Speech Translation by Understanding and Learning from the Auxiliary Text Translation Task](https://mlsanity.com/paper/2107.05782/) (2021)
- [CTC-based Compression for Direct Speech Translation](https://mlsanity.com/paper/2102.01578/) (2021)
- [RealTranS: End-to-End Simultaneous Speech Translation with Convolutional Weighted-Shrinking Transformer](https://mlsanity.com/paper/2106.04833/) (2021)
- [M-Adapter: Modality Adaptation for End-to-End Speech-to-Text Translation](https://mlsanity.com/paper/2207.00952/) (2022)
- [Pre-training for Spoken Language Understanding with Joint Textual and Phonetic Representation Learning](https://mlsanity.com/paper/2104.10357/) (2021)
- [Zero-shot Speech Translation](https://mlsanity.com/paper/2107.06010/) (2021)
- [AdaTranS: Adapting with Boundary-based Shrinking for End-to-End Speech Translation](https://mlsanity.com/paper/2212.08911/) (2022)
- [AlloST: Low-resource Speech Translation without Source Transcription](https://mlsanity.com/paper/2105.00171/) (2021)
- [RedApt: An Adaptor for wav2vec 2 Encoding  Faster and Smaller Speech Translation without Quality Compromise](https://mlsanity.com/paper/2210.08475/) (2022)
- [Bridging the Granularity Gap for Acoustic Modeling](https://mlsanity.com/paper/2305.17356/) (2023)

## Abstract (arXiv)

End-to-end models for speech translation (ST) more tightly couple speech recognition (ASR) and machine translation (MT) than a traditional cascade of separate ASR and MT models, with simpler model architectures and the potential for reduced error propagation. Their performance is often assumed to be superior, though in many conditions this is not yet the case. We compare cascaded and end-to-end models across high, medium, and low-resource conditions, and show that cascades remain stronger baselines. Further, we introduce two methods to incorporate phone features into ST models. We show that these features improve both architectures, closing the gap between end-to-end models and cascades, and outperforming previous academic work -- by up to 9 BLEU on our low-resource setting.