Optimising the Social and Environmental Sustainability of Stable Diffusion Models

Text-to-image generation models are widely used across numerous domains. Among these, Stable Diffusion (SD) — an open-source text-to-image generation model — has become the most popular, producing over 12 billion images annually. However, the widespread use of these models raises concerns regarding their social and environmental sustainability. We introduce SustainDiffusion, a search-based approach designed to enhance the social and environmental sustainability of SD models. SustainDiffusion searches the optimal combination of hyperparameters and prompt structures that reduce gender and ethnic bias in generated images while also lowering the energy consumption required for image generation. Importantly, SustainDiffusion maintains image quality comparable to that of the original SD model — demonstrating how enhancing sustainability of text-to-image generation models is possible without fine-tuning or changing the architecture.