Benchmarking Automatic Speech Recognition Models for African Languages
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908707025059840 |
|---|---|
| author | Nahabwe, Alvin Kagumire, Sulaiman Musinguzi, Denis Beijuka, Bruno Kyagaba, Jonah Mubuuke Nabende, Peter Katumba, Andrew Nakatumba-Nabende, Joyce |
| author_facet | Nahabwe, Alvin Kagumire, Sulaiman Musinguzi, Denis Beijuka, Bruno Kyagaba, Jonah Mubuuke Nabende, Peter Katumba, Andrew Nakatumba-Nabende, Joyce |
| contents | Automatic speech recognition (ASR) for African languages remains constrained by limited labeled data and the lack of systematic guidance on model selection, data scaling, and decoding strategies. Large pre-trained systems such as Whisper, XLS-R, MMS, and W2v-BERT have expanded access to ASR technology, but their comparative behavior in African low-resource contexts has not been studied in a unified and systematic way. In this work, we benchmark four state-of-the-art ASR models across 13 African languages, fine-tuning them on progressively larger subsets of transcribed data ranging from 1 to 400 hours. Beyond reporting error rates, we provide new insights into why models behave differently under varying conditions. We show that MMS and W2v-BERT are more data efficient in very low-resource regimes, XLS-R scales more effectively as additional data becomes available, and Whisper demonstrates advantages in mid-resource conditions. We also analyze where external language model decoding yields improvements and identify cases where it plateaus or introduces additional errors, depending on the alignment between acoustic and text resources. By highlighting the interaction between pre-training coverage, model architecture, dataset domain, and resource availability, this study offers practical and insights into the design of ASR systems for underrepresented languages. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_10968 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Benchmarking Automatic Speech Recognition Models for African Languages Nahabwe, Alvin Kagumire, Sulaiman Musinguzi, Denis Beijuka, Bruno Kyagaba, Jonah Mubuuke Nabende, Peter Katumba, Andrew Nakatumba-Nabende, Joyce Computation and Language Sound Audio and Speech Processing Automatic speech recognition (ASR) for African languages remains constrained by limited labeled data and the lack of systematic guidance on model selection, data scaling, and decoding strategies. Large pre-trained systems such as Whisper, XLS-R, MMS, and W2v-BERT have expanded access to ASR technology, but their comparative behavior in African low-resource contexts has not been studied in a unified and systematic way. In this work, we benchmark four state-of-the-art ASR models across 13 African languages, fine-tuning them on progressively larger subsets of transcribed data ranging from 1 to 400 hours. Beyond reporting error rates, we provide new insights into why models behave differently under varying conditions. We show that MMS and W2v-BERT are more data efficient in very low-resource regimes, XLS-R scales more effectively as additional data becomes available, and Whisper demonstrates advantages in mid-resource conditions. We also analyze where external language model decoding yields improvements and identify cases where it plateaus or introduces additional errors, depending on the alignment between acoustic and text resources. By highlighting the interaction between pre-training coverage, model architecture, dataset domain, and resource availability, this study offers practical and insights into the design of ASR systems for underrepresented languages. |
| title | Benchmarking Automatic Speech Recognition Models for African Languages |
| topic | Computation and Language Sound Audio and Speech Processing |
| url | https://arxiv.org/abs/2512.10968 |