Pith. sign in

REVIEW

Coarse-to-Fine Structure-Aware Artistic Style Transfer

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2502.05387 v1 pith:GTGMPDUW submitted 2025-02-08 cs.CV cs.AI

Coarse-to-Fine Structure-Aware Artistic Style Transfer

classification cs.CV cs.AI
keywords styleimagecontentstructuretransferlocalartisticfeatures
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Artistic style transfer aims to use a style image and a content image to synthesize a target image that retains the same artistic expression as the style image while preserving the basic content of the content image. Many recently proposed style transfer methods have a common problem; that is, they simply transfer the texture and color of the style image to the global structure of the content image. As a result, the content image has a local structure that is not similar to the local structure of the style image. In this paper, we present an effective method that can be used to transfer style patterns while fusing the local style structure into the local content structure. In our method, dif-ferent levels of coarse stylized features are first reconstructed at low resolution using a Coarse Network, in which style color distribution is roughly transferred, and the content structure is combined with the style structure. Then, the reconstructed features and the content features are adopted to synthesize high-quality structure-aware stylized images with high resolution using a Fine Network with three structural selective fusion (SSF) modules. The effectiveness of our method is demonstrated through the generation of appealing high-quality stylization results and a com-parison with some state-of-the-art style transfer methods.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.