Generative Image Compression by Estimating Gradients of the Rate-variable Feature Distribution
This work addresses the problem of generating high-quality compressed images for applications like media storage and transmission, representing an incremental improvement over prior generative compression methods.
The paper tackles generative image compression by proposing a diffusion-based framework that reinterprets compression as a forward diffusion process, achieving smooth rate adjustment and photo-realistic reconstructions with minimal sampling steps, outperforming existing methods on perceptual distortion, statistical fidelity, and no-reference quality metrics.
While learned image compression (LIC) focuses on efficient data transmission, generative image compression (GIC) extends this framework by integrating generative modeling to produce photo-realistic reconstructed images. In this paper, we propose a novel diffusion-based generative modeling framework tailored for generative image compression. Unlike prior diffusion-based approaches that indirectly exploit diffusion modeling, we reinterpret the compression process itself as a forward diffusion path governed by stochastic differential equations (SDEs). A reverse neural network is trained to reconstruct images by reversing the compression process directly, without requiring Gaussian noise initialization. This approach achieves smooth rate adjustment and photo-realistic reconstructions with only a minimal number of sampling steps. Extensive experiments on benchmark datasets demonstrate that our method outperforms existing generative image compression approaches across a range of metrics, including perceptual distortion, statistical fidelity, and no-reference quality assessments.