MLMA: Towards Multilingual ASR With Mamba-based Architectures
This addresses the challenge of balancing performance across high- and low-resource languages in multilingual ASR, though it appears incremental as it adapts an existing architecture to this domain.
The paper tackles multilingual automatic speech recognition by introducing MLMA, a Mamba-based architecture that achieves competitive performance compared to Transformer-based methods on standard benchmarks.
Multilingual automatic speech recognition (ASR) remains a challenging task, especially when balancing performance across high- and low-resource languages. Recent advances in sequence modeling suggest that architectures beyond Transformers may offer better scalability and efficiency. In this work, we introduce MLMA (Multilingual Language Modeling with Mamba for ASR), a new approach that leverages the Mamba architecture -- an efficient state-space model optimized for long-context sequence processing -- for multilingual ASR. Using Mamba, MLMA implicitly incorporates language-aware conditioning and shared representations to support robust recognition across diverse languages. Experiments on standard multilingual benchmarks show that MLMA achieves competitive performance compared to Transformer-based architectures. These results highlight Mamba's potential as a strong backbone for scalable, efficient, and accurate multilingual speech recognition.