Diffusion Models Through a Global Lens: Are They Culturally Inclusive?
This work addresses the problem of cultural inclusivity in AI-generated images, which affects the global community, particularly underrepresented cultures, by highlighting the need for more inclusive generative AI systems.
The study evaluated state-of-the-art diffusion models on their ability to generate culturally specific images, finding significant disparities in cultural relevance, description fidelity, and realism, with models often failing to accurately represent cultural nuances across ten countries. The results showed notable shortcomings, particularly for underrepresented regions.
Text-to-image diffusion models have recently enabled the creation of visually compelling, detailed images from textual prompts. However, their ability to accurately represent various cultural nuances remains an open question. In our work, we introduce CultDiff benchmark, evaluating state-of-the-art diffusion models whether they can generate culturally specific images spanning ten countries. We show that these models often fail to generate cultural artifacts in architecture, clothing, and food, especially for underrepresented country regions, by conducting a fine-grained analysis of different similarity aspects, revealing significant disparities in cultural relevance, description fidelity, and realism compared to real-world reference images. With the collected human evaluations, we develop a neural-based image-image similarity metric, namely, CultDiff-S, to predict human judgment on real and generated images with cultural artifacts. Our work highlights the need for more inclusive generative AI systems and equitable dataset representation over a wide range of cultures.