Towards Extreme Image Compression with Latent Feature Guidance and Diffusion Prior
This addresses the challenge of high-quality image compression at very low bitrates for applications like storage and transmission, representing an incremental advance in leveraging generative models.
The paper tackles extreme image compression at bitrates below 0.1 bpp by proposing a two-stage framework that uses pre-trained diffusion models for realistic reconstruction, achieving significant visual performance improvements over state-of-the-art methods.
Image compression at extremely low bitrates (below 0.1 bits per pixel (bpp)) is a significant challenge due to substantial information loss. In this work, we propose a novel two-stage extreme image compression framework that exploits the powerful generative capability of pre-trained diffusion models to achieve realistic image reconstruction at extremely low bitrates. In the first stage, we treat the latent representation of images in the diffusion space as guidance, employing a VAE-based compression approach to compress images and initially decode the compressed information into content variables. The second stage leverages pre-trained stable diffusion to reconstruct images under the guidance of content variables. Specifically, we introduce a small control module to inject content information while keeping the stable diffusion model fixed to maintain its generative capability. Furthermore, we design a space alignment loss to force the content variables to align with the diffusion space and provide the necessary constraints for optimization. Extensive experiments demonstrate that our method significantly outperforms state-of-the-art approaches in terms of visual performance at extremely low bitrates. The source code and trained models are available at https://github.com/huai-chang/DiffEIC.