Vision-Centric Data Augmentation Techniques for Language Models. IJCHML 2025, 3 (3).