Datasets
Standard Dataset
StyleBench
- Citation Author(s):
- Submitted by:
- CaiRong Zhao
- Last updated:
- Mon, 02/24/2025 - 07:15
- DOI:
- 10.21227/t070-v756
- License:
6 Views- Categories:
- Keywords:
Abstract
Style Transfer Evaluation Benchmark (StyleBench) in submitted paper "StyleShot: A SnapShot on any Style".
To comprehensively evaluate the effectiveness and generalization ability of style transfer methods, we build StyleBench that covers 73 distinct styles, ranging from paintings, flat illustrations, 3D rendering to sculptures with varying materials. For each style, we collect 5-7 distinct images with variations.
In total, our StyleBench contains 490 images across diverse styles.
Moreover, we generated 40 content images from simple to complex that describe random objects and scenarios as content input.
Our StyleBench consists of two dictionaries: "styles" and "content."
The "styles" dictionary contains 73 styles, with each style having 5-7 images in the corresponding folder, such as "3D Model," "Anime," "Pop Art," and so on.
The "content" dictionary includes 40 content images.
