FreeU: Free Lunch in Diffusion U-Net
This provides a simple, training-free improvement for diffusion models like Stable Diffusion, benefiting users in image and video generation, though it is incremental as it builds on existing architectures.
The paper tackles the issue of diffusion U-Net architectures overlooking backbone semantics due to skip connections, and proposes FreeU, a method that re-weights skip connections and backbone features to improve generation quality without extra training, achieving enhanced results in image and video generation tasks.
In this paper, we uncover the untapped potential of diffusion U-Net, which serves as a "free lunch" that substantially improves the generation quality on the fly. We initially investigate the key contributions of the U-Net architecture to the denoising process and identify that its main backbone primarily contributes to denoising, whereas its skip connections mainly introduce high-frequency features into the decoder module, causing the network to overlook the backbone semantics. Capitalizing on this discovery, we propose a simple yet effective method-termed "FreeU" - that enhances generation quality without additional training or finetuning. Our key insight is to strategically re-weight the contributions sourced from the U-Net's skip connections and backbone feature maps, to leverage the strengths of both components of the U-Net architecture. Promising results on image and video generation tasks demonstrate that our FreeU can be readily integrated to existing diffusion models, e.g., Stable Diffusion, DreamBooth, ModelScope, Rerender and ReVersion, to improve the generation quality with only a few lines of code. All you need is to adjust two scaling factors during inference. Project page: https://chenyangsi.top/FreeU/.