RDEIC: Accelerating Diffusion-Based Extreme Image Compression With Relay Residual Diffusion
- 9 June 2025
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Video Technology
- Vol. 35 (11) , 11540-11552
Abstract
Diffusion-based extreme image compression methods have achieved impressive performance at extremely low bitrates. However, constrained by the iterative denoising process that starts from pure noise, these methods are limited in both fidelity and efficiency. To address these two issues, we present Relay Residual Diffusion Extreme Image Compression (RDEIC), which leverages compressed feature initialization and residual diffusion. Specifically, we first use the compressed latent features of the image with added noise, instead of pure noise, as the starting point to eliminate the unnecessary initial stages of the denoising process. Second, we directly derive a novel residual diffusion equation from Stable Diffusion’s original diffusion equation that reconstructs the raw image by iteratively removing the added noise and the residual between the compressed and target latent features. In this way, we effectively combine the efficiency of residual diffusion with the powerful generative capability of Stable Diffusion. Third, we propose a fixed-step fine-tuning strategy to eliminate the discrepancy between the training and inference phases, thereby further improving the reconstruction quality. Extensive experiments demonstrate that the proposed RDEIC achieves state-of-the-art visual quality and outperforms existing diffusion-based extreme image compression methods in both fidelity and efficiency. The source code and pre-trained models are available at https://github.com/huai-chang/RDEIC.Keywords
Funding Information
- National Natural Science Foundation of China (NSFC62088102, NSFC62376208, U24A20265)
- Key Core Technology Research Project in Shaanxi Province (2024SF-GJHX-27)
- Qin Chuangyuan General Window "Four Chain" Integration Project (2024PT-ZCK-66)
This publication has 26 references indexed in Scilit:
- Exploiting Diffusion Prior for Real-World Image Super-ResolutionInternational Journal of Computer Vision, 2024
- Generative Latent Coding for Ultra-Low Bitrate Image Compression2024
- SeeSR: Towards Semantics-Aware Real-World Image Super-Resolution2024
- Learned Image Compression with Mixed Transformer-CNN Architectures2023
- LSDIR: A Large Scale Dataset for Image Restoration2023
- ELIC: Efficient Learned Image Compression with Unevenly Grouped Space-Channel Contextual Adaptive Coding2022
- Checkerboard Context Model for Efficient Learned Image Compression2021
- Deep Homography for Efficient Stereo Image Compression2021
- The Unreasonable Effectiveness of Deep Features as a Perceptual Metric2018
- Multiscale structural similarity for image quality assessment2004