Annotation-Efficient Vision-Language Model Adaptation to the Polish Language Using the LLaVA Framework
This work addresses the limited usability of vision-language models for non-English-speaking users, specifically for Polish, though it is incremental as it adapts an existing methodology.
The authors tackled the problem of adapting vision-language models to the Polish language using automated translation and synthetic data, resulting in a +9.5% improvement over a baseline model on a Polish-adapted benchmark and higher-quality captions in human evaluations.
Most vision-language models (VLMs) are trained on English-centric data, limiting their performance in other languages and cultural contexts. This restricts their usability for non-English-speaking users and hinders the development of multimodal systems that reflect diverse linguistic and cultural realities. In this work, we reproduce and adapt the LLaVA-Next methodology to create a set of Polish VLMs. We rely on a fully automated pipeline for translating and filtering existing multimodal datasets, and complement this with synthetic Polish data for OCR and culturally specific tasks. Despite relying almost entirely on automatic translation and minimal manual intervention to the training data, our approach yields strong results: we observe a +9.5% improvement over LLaVA-1.6-Vicuna-13B on a Polish-adapted MMBench, along with higher-quality captions in generative evaluations, as measured by human annotators in terms of linguistic correctness. These findings highlight that large-scale automated translation, combined with lightweight filtering, can effectively bootstrap high-quality multimodal models for low-resource languages. Some challenges remain, particularly in cultural coverage and evaluation. To facilitate further research, we make our models and evaluation dataset publicly available.