English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Recent years have witnessed the remarkable performance of diffusion models invarious vision tasks. However, for image restoration that aims to recover clearimages with sharper details from given degraded observations, diffusion-basedmethods may fail to recover promising results due to inaccurate noiseestimation. Moreover, simple constraining noises cannot effectively learncomplex degradation information, which subsequently hinders the model capacity.To solve the above problems, we propose a coarse-to-fine diffusion Transformer(C2F-DFT) for image restoration. Specifically, our C2F-DFT contains diffusionself-attention (DFSA) and diffusion feed-forward network (DFN) within a newcoarse-to-fine training scheme. The DFSA and DFN respectively capture thelong-range diffusion dependencies and learn hierarchy diffusion representationto facilitate better restoration. In the coarse training stage, our C2F-DFTestimates noises and then generates the final clean image by a samplingalgorithm. To further improve the restoration quality, we propose a simple yeteffective fine training scheme. It first exploits the coarse-trained diffusionmodel with fixed steps to generate restoration results, which then would beconstrained with corresponding ground-truth ones to optimize the models toremedy the unsatisfactory results affected by inaccurate noise estimation.Extensive experiments show that C2F-DFT significantly outperformsdiffusion-based restoration method IR-SDE and achieves competitive performancecompared with Transformer-based state-of-the-art methods on $3$ tasks,including image deraining, image deblurring, and real image denoising. Code isavailable at https://github.com/wlydlut/C2F-DFT.

Learning A Coarse-to-Fine Diffusion Transformer for Image Restoration