# Speech translation

Area: [Speech and audio language models](https://mlsanity.com/speech/index.md). HTML: https://mlsanity.com/speech/speech-translation/

## Key facts

- Speech translation: 365 papers on arXiv since 2021 (data through 19 Oct 2026).
- 70 of its 365 mapped papers (19%) cite “Direct speech-to-speech translation with discrete units” (2021, arXiv 2107.05604).
- 55 of its 365 mapped papers (15%) cite “Translatotron 2: High-quality direct speech-to-speech translation with voice preservation” (2021, arXiv 2107.08661).
- 4 new papers 2 Oct–19 Oct.

## Most cited here

- [Direct speech-to-speech translation with discrete units](https://mlsanity.com/paper/2107.05604/) (2021), cited by 70 papers here
- [Translatotron 2: High-quality direct speech-to-speech translation with voice preservation](https://mlsanity.com/paper/2107.08661/) (2021), cited by 55 papers here
- [Textless Speech-to-Speech Translation on Real Data](https://mlsanity.com/paper/2112.08352/) (2021), cited by 50 papers here
- [CVSS Corpus and Massively Multilingual Speech-to-Speech Translation](https://mlsanity.com/paper/2201.03713/) (2022), cited by 44 papers here
- [UnitY: Two-pass Direct Speech-to-speech Translation with Discrete Units](https://mlsanity.com/paper/2212.08055/) (2022), cited by 40 papers here
- [Improving Speech Translation by Understanding and Learning from the Auxiliary Text Translation Task](https://mlsanity.com/paper/2107.05782/) (2021), cited by 37 papers here
- [Cascade versus Direct Speech Translation: Do the Differences Still Make a Difference?](https://mlsanity.com/paper/2106.01045/) (2021), cited by 35 papers here
- [End-to-end Speech Translation via Cross-modal Progressive Training](https://mlsanity.com/paper/2104.10380/) (2021), cited by 32 papers here
- [Stacked Acoustic-and-Textual Encoding: Integrating the Pre-trained Models into Speech Translation Encoders](https://mlsanity.com/paper/2105.05752/) (2021), cited by 32 papers here
- [STEMM: Self-learning with Speech-text Manifold Mixup for Speech Translation](https://mlsanity.com/paper/2203.10426/) (2022), cited by 32 papers here

## New (2 Oct–19 Oct)

- [Cross-Lingual TTS for Indian Languages: A Benchmark and Inference-Time Duration Estimation Strategies](https://mlsanity.com/paper/2610.14826/) (2026-10-10)
- [Dialect-Robust Speech Language Models with Synthetic Pseudo-Dialect Augmentation](https://mlsanity.com/paper/2610.09321/) (2026-10-07)
- [Quality-Aware Self-Correcting Speech Translation on an Edge Device](https://mlsanity.com/paper/2610.07545/) (2026-10-06)
- [Learning When to Commit from Partial Speech for End-to-End Simultaneous Speech Translation](https://mlsanity.com/paper/2610.02612/) (2026-10-02)