# Replay to Remember: Continual Layer-Specific Fine-tuning for German Speech Recognition

arXiv: https://arxiv.org/abs/2307.07280 · HTML: https://mlsanity.com/paper/2307.07280/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Builds on 15 of its 22 references that are in the map.
- First posted on arXiv 14 Jul 2023.

## Builds on (most cited first, 15 of 15)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Robust Speech Recognition via Large-Scale Weak Supervision](https://mlsanity.com/paper/2212.04356/) (2022)
- [wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations](https://mlsanity.com/paper/2006.11477/) (2020)
- [Conformer: Convolution-augmented Transformer for Speech Recognition](https://mlsanity.com/paper/2005.08100/) (2020)
- [Common Voice: A Massively-Multilingual Speech Corpus](https://mlsanity.com/paper/1912.06670/) (2019)
- [Continual Lifelong Learning with Neural Networks: A Review](https://mlsanity.com/paper/1802.07569/) (2018)
- [Experience Replay for Continual Learning](https://mlsanity.com/paper/1811.11682/) (2018)
- [XLS-R: Self-supervised Cross-lingual Speech Representation Learning at Scale](https://mlsanity.com/paper/2111.09296/) (2021)
- [Listen, Attend and Spell](https://mlsanity.com/paper/1508.01211/) (2015)
- [Unsupervised Cross-lingual Representation Learning for Speech Recognition](https://mlsanity.com/paper/2006.13979/) (2020)
- [Personalizing ASR for Dysarthric and Accented Speech with Limited Data](https://mlsanity.com/paper/1907.13511/) (2019)
- [Conv-Transformer Transducer: Low Latency, Low Frame Rate, Streamable End-to-End Speech Recognition](https://mlsanity.com/paper/2008.05750/) (2020)
- [Hey ASR System! Why Aren't You More Inclusive? Automatic Speech Recognition Systems' Bias and Proposed Bias Mitigation Techniques. A Literature Review](https://mlsanity.com/paper/2211.09511/) (2022)
- [Echo State Speech Recognition](https://mlsanity.com/paper/2102.09114/) (2021)
- [Continual Learning for Monolingual End-to-End Automatic Speech Recognition](https://mlsanity.com/paper/2112.09427/) (2021)

## Built on by (most cited first, 4 of 4)

- [Large Language Model Data Generation for Enhanced Intent Recognition in German Speech](https://mlsanity.com/paper/2508.06277/) (2025)
- [End-to-End Transformer-based Automatic Speech Recognition for Northern Kurdish: A Pioneering Approach](https://mlsanity.com/paper/2410.16330/) (2024)
- [Learning to Hear Hesitation: Continual Learning for Disfluency-Aware ASR](https://mlsanity.com/paper/2606.14391/) (2026)
- [Unified Gradient Projection: Language-Balanced Continual Learning for Multilingual Low-Resource ASR](https://mlsanity.com/paper/2607.11163/) (2026)

## Abstract (arXiv)

While Automatic Speech Recognition (ASR) models have shown significant advances with the introduction of unsupervised or self-supervised training techniques, these improvements are still only limited to a subsection of languages and speakers. Transfer learning enables the adaptation of large-scale multilingual models to not only low-resource languages but also to more specific speaker groups. However, fine-tuning on data from new domains is usually accompanied by a decrease in performance on the original domain. Therefore, in our experiments, we examine how well the performance of large-scale ASR models can be approximated for smaller domains, with our own dataset of German Senior Voice Commands (SVC-de), and how much of the general speech recognition performance can be preserved by selectively freezing parts of the model during training. To further increase the robustness of the ASR model to vocabulary and speakers outside of the fine-tuned domain, we apply Experience Replay for continual learning. By adding only a fraction of data from the original domain, we are able to reach Word-Error-Rates (WERs) below 5% on the new domain, while stabilizing performance for general speech recognition at acceptable WERs.