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 \textbf{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 \textbf{sub-second} inference speed for generating a 512×512 image on mobile devices, establishing a new state of the art.
@article{arxiv.2311.16567,
title = {MobileDiffusion: Instant Text-to-Image Generation on Mobile Devices},
author = {Yang Zhao and Yanwu Xu and Zhisheng Xiao and Haolin Jia and Tingbo Hou},
journal= {arXiv preprint arXiv:2311.16567},
year = {2024}
}