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 Plus annual

Presents the access of premium content as premium feature

Premium Content

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 Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Recent work has showcased the significant potential of diffusion models inpose-guided person image synthesis. However, owing to the inconsistency in posebetween the source and target images, synthesizing an image with a distinctpose, relying exclusively on the source image and target pose information,remains a formidable challenge. This paper presents Progressive ConditionalDiffusion Models (PCDMs) that incrementally bridge the gap between personimages under the target and source poses through three stages. Specifically, inthe first stage, we design a simple prior conditional diffusion model thatpredicts the global features of the target image by mining the global alignmentrelationship between pose coordinates and image appearance. Then, the secondstage establishes a dense correspondence between the source and target imagesusing the global features from the previous stage, and an inpaintingconditional diffusion model is proposed to further align and enhance thecontextual features, generating a coarse-grained person image. In the thirdstage, we propose a refining conditional diffusion model to utilize thecoarsely generated image from the previous stage as a condition, achievingtexture restoration and enhancing fine-detail consistency. The three-stagePCDMs work progressively to generate the final high-quality and high-fidelitysynthesized image. Both qualitative and quantitative results demonstrate theconsistency and photorealism of our proposed PCDMs under challengingscenarios.The code and model will be available athttps://github.com/muzishen/PCDMs.

Advancing Pose-Guided Image Synthesis with Progressive Conditional Diffusion Models