Papers for

mobile assistant developers

Papers whose findings have a practical use for this group, as judged from the abstract. Open a paper to read what it means in practice.

Multilingual speech models improve low-resource languages with adapter stacking

Sequential Adapter Stacking for Cross-Lingual Low-Resource ASR

Abstract: Extending large-scale multilingual automatic speech recognition (ASR) models to low-resource languages remains challenging. Model performance is skewed toward high-resource languages and degrades sharply for languages with limited labeled data and pre-training exposure. To address this, we investigate parameter-efficient approaches for transferring knowledge from resource-rich source languages to low-resource target languages on Whisper. Alongside warm initialization and attention-based fusion, we propose Sequential Adapter Stacking, which places a trainable target-language adapter on top of a frozen source-language adapter. Under controlled experiments, these approaches are evaluated on three target languages unsupported by Whisper -- Asturian, Assamese, and Xhosa -- using source languages with varying degrees of relatedness. Sequential Adapter Stacking with the closest related source consistently and significantly outperforms full fine-tuning across the three targets, with 5--8\% relative WER reductions. These gains largely persist with only one hour of target training data.

Mon 14 SeptComputation and Language
The gist
Speech recognition models often work well for languages with lots of data but poorly for languages with very little training data. The authors study ways to improve these models for low-resource languages by adding small extra parts called adapters that learn from related languages. They test their approach on three languages not supported before and find that stacking adapters in sequence works better than retraining the whole model. This method still helps even when only one hour of new language data is available.
Open 2609.15758v1