# From Monolingual to Multilingual: Evaluating Mamba for ASR in South African Languages

arXiv: https://arxiv.org/abs/2607.01502 · HTML: https://mlsanity.com/paper/2607.01502/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Speech and audio SSMs, one of 10 directions in State-space sequence models (S4, Mamba).
- Builds on 24 of its 33 references that are in the map.
- First posted on arXiv 1 Jul 2026.

## Placed in

- State-space sequence models (S4, Mamba) / [Speech and audio SSMs](https://mlsanity.com/ssm/speech-and-audio-ssms/)

## Builds on (most cited first, 24 of 24)

- [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://mlsanity.com/paper/2312.00752/) (2023)
- [Robust Speech Recognition via Large-Scale Weak Supervision](https://mlsanity.com/paper/2212.04356/) (2022)
- [Conformer: Convolution-augmented Transformer for Speech Recognition](https://mlsanity.com/paper/2005.08100/) (2020)
- [Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context](https://mlsanity.com/paper/1901.02860/) (2019)
- [Common Voice: A Massively-Multilingual Speech Corpus](https://mlsanity.com/paper/1912.06670/) (2019)
- [FLEURS: Few-shot Learning Evaluation of Universal Representations of Speech](https://mlsanity.com/paper/2205.12446/) (2022)
- [End-to-End Speech Recognition: A Survey](https://mlsanity.com/paper/2303.03329/) (2023)
- [E-Branchformer: Branchformer with Enhanced merging for speech recognition](https://mlsanity.com/paper/2210.00077/) (2022)
- [mHuBERT-147: A Compact Multilingual HuBERT Model](https://mlsanity.com/paper/2406.06371/) (2024)
- [The Low-Resource Double Bind: An Empirical Study of Pruning for Low-Resource Machine Translation](https://mlsanity.com/paper/2110.03036/) (2021)
- [Speech Slytherin: Examining the Performance and Efficiency of Mamba for Speech Separation, Recognition, and Synthesis](https://mlsanity.com/paper/2407.09732/) (2024)
- [LAE: Language-Aware Encoder for Monolingual and Multilingual ASR](https://mlsanity.com/paper/2206.02093/) (2022)
- [Exploring the Capability of Mamba in Speech Applications](https://mlsanity.com/paper/2406.16808/) (2024)
- [AfriHuBERT: A self-supervised speech representation model for African languages](https://mlsanity.com/paper/2409.20201/) (2024)
- [Charting the Landscape of African NLP: Mapping Progress and Shaping the Road Ahead](https://mlsanity.com/paper/2505.21315/) (2025)
- [Building a Unified Code-Switching ASR System for South African Languages](https://mlsanity.com/paper/1807.10949/) (2018)
- [WAXAL: A Large-Scale Multilingual African Language Speech Corpus](https://mlsanity.com/paper/2602.02734/) (2026)
- [Semi-supervised acoustic model training for five-lingual code-switched ASR](https://mlsanity.com/paper/1906.08647/) (2019)
- [Semi-supervised Development of ASR Systems for Multilingual Code-switched Speech in Under-resourced Languages](https://mlsanity.com/paper/2003.03135/) (2020)
- [Automatic Speech Recognition for African Low-Resource Languages: Challenges and Future Directions](https://mlsanity.com/paper/2505.11690/) (2025)
- [The Esethu Framework: Reimagining Sustainable Dataset Governance and Curation for Low-Resource Languages](https://mlsanity.com/paper/2502.15916/) (2025)
- [Swivuriso: The South African Next Voices Multilingual Speech Dataset](https://mlsanity.com/paper/2512.02201/) (2025)
- [Beyond the Utterance: An Empirical Study of Very Long Context Speech Recognition](https://mlsanity.com/paper/2602.09044/) (2026)
- [MLMA: Towards Multilingual ASR With Mamba-based Architectures](https://mlsanity.com/paper/2510.18684/) (2025)

## Abstract (arXiv)

Recent advances in automatic speech recognition (ASR) have explored different sequence models, including Conformer-based models and newer state space models such as Mamba. Although prior work has evaluated these architectures in multiple languages, their effectiveness in African languages remains underexplored. In this work, we evaluate Mamba for ASR on seven South African languages. In monolingual experiments, each model is trained on 50 hours of speech per language, and we compare Mamba to a Conformer baseline of similar parameter scale. Mamba achieves similar recognition accuracy to Conformer while using fewer computational resources and training faster. We further evaluate generalization in this setting and find that both models struggle to generalize to speech that is much longer than what they were trained on. We then study multilingual ASR using Mamba models, where the baseline is pooling all languages together. On top of this, we tested three extensions: training with language-family information by adding both language and language-family embeddings as biases to the downsampled acoustic representations, and multitask learning with a CTC ASR objective and a language identification (LID) head. We find that multilingual training consistently improves performance over monolingual training. However, adding explicit language information does not improve in-domain performance but does improve cross-corpus robustness. We conducted ablation studies in low-resource multilingual settings using 5-hour and 10-hour per-language training data, where we observed gains from using language embeddings and further demonstrated that removing or altering them hurt model performance. Lastly, we analysed these embeddings and find that they do not capture linguistic similarity in a typological sense, but instead act as task-specific control vectors.