Inf-DiT: Upsampling Any-Resolution Image with Memory-Efficient Diffusion Transformer
This addresses a critical bottleneck for researchers and practitioners in high-resolution image generation, enabling more efficient and scalable applications.
The paper tackles the memory inefficiency of generating ultra-high-resolution images (e.g., 4096x4096) with diffusion models by proposing a unidirectional block attention mechanism, achieving state-of-the-art performance and saving over 5x memory compared to UNet structures.
Diffusion models have shown remarkable performance in image generation in recent years. However, due to a quadratic increase in memory during generating ultra-high-resolution images (e.g. 4096*4096), the resolution of generated images is often limited to 1024*1024. In this work. we propose a unidirectional block attention mechanism that can adaptively adjust the memory overhead during the inference process and handle global dependencies. Building on this module, we adopt the DiT structure for upsampling and develop an infinite super-resolution model capable of upsampling images of various shapes and resolutions. Comprehensive experiments show that our model achieves SOTA performance in generating ultra-high-resolution images in both machine and human evaluation. Compared to commonly used UNet structures, our model can save more than 5x memory when generating 4096*4096 images. The project URL is https://github.com/THUDM/Inf-DiT.