SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2303.13511 · CVPR · 2023

Neural Preset for Color Style Transfer

Nanxuan Zhao, Rynson Lau, Lei Zhu, Zhanghan Ke, Yuhao Liu

arXiv · PDF · Open in the Atlas

Code that ran

We have not lifted any functions out of this paper's repositories yet, so there is nothing we have run. If it links a repository, it is listed below.

Abstract

Adobe Research Figure 1. Our Color Style Transfer Results. (a) State-of-the-art methods PhotoNAS [1] and PhotoWCT 2 [6] produce distort textures (e.g., text in green box) and dissonant colors (e.g., content in orange box). Besides, they have long inference time even on the latest Nvidia RTX3090 GPU (red numbers in brackets). In contrast, our method avoids artifacts and is ∼28× faster. (b) Our method can produce faithful results on 8K images, but both PhotoNAS and PhotoWCT 2 run into the out-of-memory problem. Zoom in for better visualization.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2303.13511")
get_code_for_paper("2303.13511")
have("2303.13511")

Connect an agent — have() is free.