Evaluating the Effect of Linguistic Relatedness on Cross-Lingual Transfer in Large Multilingual Automatic Speech Recognition
2026-07-06 • Computation and Language
Computation and LanguageArtificial Intelligence
AI summaryⓘ
The authors studied how to improve speech recognition for African languages that have very little training data. They tested if using related languages first could help big speech recognition models learn better. Their experiments showed that, unlike small models, large models did not benefit much from this approach when there was only a little target language data. This means that just picking a related language might not be enough to help big speech models work better on low-resource languages.
automatic speech recognitionlow-resource languagescross-lingual transfermultilingual ASRlinguistic relatednesslanguage adaptationAfrican languagestransfer learningdeep learning models
Authors
Andrei Florian, Cynthia Jayne Amol, Hope Kerubo Ombaba, Xiaoyu Cui, Boniface Mwau, Biatus Maina Kamau, Lilian Diana Awuor Wanzare, Christiane Fellbaum, Happy Buzaaba
Abstract
Extending automatic speech recognition (ASR) to low-resource African languages is constrained by the prohibitive demands of data collection at scale. A promising direction is to leverage linguistic relatedness to enhance cross-lingual transfer from a related auxiliary language to the low-resource target by sequentially adapting on both. Although this strategy has shown meaningful improvements in small ASR models, its effectiveness in large ASR remains unclear. We extend this framework to large multilingual ASR through a systematic controlled experimental design spanning six factors, two Africa-centric corpora, and four large ASR models, isolating whether linguistic relatedness reliably predicts cross-lingual transfer gains in this setting. Across all conditions, pre-adaptation on related auxiliary languages yields no practically meaningful transfer improvements given minimal target-language data, suggesting that linguistic relatedness alone may not reliably predict cross-lingual transfer gains in large multilingual ASR, or constitute an effective strategy for extending such models to low-resource languages.