What question did this study set out to answer?

This research aims to improve image reconstruction quality under low bitrate conditions using diffusion models effectively.

DiRIC: Diffusion Prior Refinement for Efficient Low-rate Image Compression

View Full Paper

Ask AI

Bookmark

Authors

Randomized trial demonstrates enhanced image compression fidelity in low-rate conditions, suggesting improved decoding efficiency.

This research aims to improve image reconstruction quality under low bitrate conditions using diffusion models effectively.
Proposed the DiRIC compression scheme based on Stable Diffusion.
Employed a multi-feature compressor to represent low-level information efficiently.
Integrated a pre-embedding mechanism and a noise-level estimator for enhanced decoding.
Achieved a 147.44% improvement in FID and an 84.63% improvement in PSNR compared to previous diffusion baselines.
Demonstrated a 19× increase in decoding speed.
Outperformed prior perceptual image compression methods in terms of statistical fidelity.

Xia et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark