MobileDiffusion: Instant Text-to-Image Generation on Mobile Devices

The deployment of large-scale text-to-image diffusion models on mobile devices is impeded by their substantial model size and slow inference speed. In this paper, we propose $\mathbf{\text{MobileDiffusion}}$, a highly efficient text-to-image diffusion model obtained through extensive optimizations in both architecture and sampling techniques. We conduct a comprehensive examination of model architecture design to reduce redundancy, enhance computational efficiency, and minimize model’s parameter count, while preserving image generation quality. Additionally, we employ distillation and diffusion-GAN finetuning techniques on MobileDiffusion to achieve 8-step and 1-step inference respectively. Empirical studies, conducted both quantitatively and qualitatively, demonstrate the effectiveness of our proposed techniques. MobileDiffusion achieves a remarkable $\mathbf{\text{sub-second}}$ inference speed for generating a $512\times 512$ image on mobile devices, establishing a new state of the art.