Diverse Diffusion: Enhancing Image Diversity in Text-to-Image Generation
This addresses the problem of generating more inclusive and varied AI art for users of text-to-image systems, though it is incremental as it builds on existing models like Stable Diffusion.
The paper tackles the lack of diversity in images generated by text-to-image models by introducing Diverse Diffusion, an unsupervised method that finds distant vectors in the latent space to enhance diversity in aspects like color and ethnicity/gender representation, resulting in improved metrics such as LPIPS and color diversity.
Latent diffusion models excel at producing high-quality images from text. Yet, concerns appear about the lack of diversity in the generated imagery. To tackle this, we introduce Diverse Diffusion, a method for boosting image diversity beyond gender and ethnicity, spanning into richer realms, including color diversity.Diverse Diffusion is a general unsupervised technique that can be applied to existing text-to-image models. Our approach focuses on finding vectors in the Stable Diffusion latent space that are distant from each other. We generate multiple vectors in the latent space until we find a set of vectors that meets the desired distance requirements and the required batch size.To evaluate the effectiveness of our diversity methods, we conduct experiments examining various characteristics, including color diversity, LPIPS metric, and ethnicity/gender representation in images featuring humans.The results of our experiments emphasize the significance of diversity in generating realistic and varied images, offering valuable insights for improving text-to-image models. Through the enhancement of image diversity, our approach contributes to the creation of more inclusive and representative AI-generated art.