AI RESEARCH
DynT2I-Eval: A Dynamic Evaluation Framework for Text-to-Image Models
arXiv CS.CV
•
ArXi:2605.06170v1 Announce Type: new Existing text-to-image (T2I) benchmarks largely rely on fixed prompt sets, leaving them vulnerable to overfitting and benchmark contamination once publicly released and repeatedly reused. In this work, we propose DynT2I-Eval, a fully automated dynamic evaluation framework for T2I models. It constructs a structured visual semantic space from long-form descriptions, decomposing prompts into controllable dimensions (e.g., subject, logical constraint, environment, and composition.