# Comparison of Self-Supervised Speech Pre-Training Methods on Flemish Dutch

arXiv: https://arxiv.org/abs/2109.14357 · HTML: https://mlsanity.com/paper/2109.14357/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 29 Sep 2021.

## Built on by (most cited first, 2 of 2)

- [Learning to Jointly Transcribe and Subtitle for End-to-End Spontaneous Speech Recognition](https://mlsanity.com/paper/2210.07771/) (2022)
- [On the Cross-lingual Transferability of Pre-trained wav2vec2-based Models](https://mlsanity.com/paper/2511.21704/) (2025)

## Abstract (arXiv)

Recent research in speech processing exhibits a growing interest in unsupervised and self-supervised representation learning from unlabelled data to alleviate the need for large amounts of annotated data. We investigate several popular pre-training methods and apply them to Flemish Dutch. We compare off-the-shelf English pre-trained models to models trained on an increasing amount of Flemish data. We find that the most important factors for positive transfer to downstream speech recognition tasks include a substantial amount of data and a matching pre-training domain. Ideally, we also finetune on an annotated subset in the target language. All pre-trained models improve linear phone separability in Flemish, but not all methods improve Automatic Speech Recognition. We experience superior performance with wav2vec 2.0 and we obtain a 30% WER improvement by finetuning the multilingually pre-trained XLSR-53 model on Flemish Dutch, after integration into an HMM-DNN acoustic model.