Role-SynthCLIP: A Role Play Driven Diverse Synthetic Data Approach

Yuanxiang Huangfu, Chaochao Wang, Weilei Wang

The effectiveness of Contrastive Language-Image Pre-training (CLIP) models critically depends on the semantic diversity and quality of their training data. However, while existing…