One-Step Diffusion for Perceptual Image Compression
This work addresses a practical bottleneck for deploying diffusion-based compression in real-world applications, though it is incremental as it builds on existing diffusion methods.
The paper tackles the high inference latency and computational overhead in diffusion-based image compression by proposing a method that uses only a single-step diffusion process, achieving a 46× faster inference speed while maintaining comparable compression performance.
Diffusion-based image compression methods have achieved notable progress, delivering high perceptual quality at low bitrates. However, their practical deployment is hindered by significant inference latency and heavy computational overhead, primarily due to the large number of denoising steps required during decoding. To address this problem, we propose a diffusion-based image compression method that requires only a single-step diffusion process, significantly improving inference speed. To enhance the perceptual quality of reconstructed images, we introduce a discriminator that operates on compact feature representations instead of raw pixels, leveraging the fact that features better capture high-level texture and structural details. Experimental results show that our method delivers comparable compression performance while offering a 46$\times$ faster inference speed compared to recent diffusion-based approaches. The source code and models are available at https://github.com/cheesejiang/OSDiff.