Where Culture Fades: Revealing the Cultural Gap in Text-to-Image Generation
This addresses the cultural gap in text-to-image models, which is an incremental improvement for users needing culturally accurate image synthesis.
The paper tackles the problem of cultural inconsistency in multilingual text-to-image generation, showing that current models produce culturally neutral or English-biased outputs. It proposes a probing method and two alignment strategies, achieving consistent improvements in cultural consistency on CultureBench while preserving fidelity and diversity.
Multilingual text-to-image (T2I) models have advanced rapidly in terms of visual realism and semantic alignment, and are now widely utilized. Yet outputs vary across cultural contexts: because language carries cultural connotations, images synthesized from multilingual prompts should preserve cross-lingual cultural consistency. We conduct a comprehensive analysis showing that current T2I models often produce culturally neutral or English-biased results under multilingual prompts. Analyses of two representative models indicate that the issue stems not from missing cultural knowledge but from insufficient activation of culture-related representations. We propose a probing method that localizes culture-sensitive signals to a small set of neurons in a few fixed layers. Guided by this finding, we introduce two complementary alignment strategies: (1) inference-time cultural activation that amplifies the identified neurons without backbone fine-tuned; and (2) layer-targeted cultural enhancement that updates only culturally relevant layers. Experiments on our CultureBench demonstrate consistent improvements over strong baselines in cultural consistency while preserving fidelity and diversity.