arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2510.16319cs.CV
收藏

Stroke2Sketch:利用笔触属性实现免训练草图生成

Stroke2Sketch: Harnessing Stroke Attributes for Training-Free Sketch Generation

Huaqiao University(华侨大学);South China University of Technology(华南理工大学);Shaanxi Normal University(陕西师范大学);Beijing University of Aeronautics and Astronautics(北京航空航天大学);Singapore Management University(新加坡国立大学)

Rui Yang, Huining Li, Yiyi Long, Xiaojun Wu, Shengfeng He

AI总结:

本文提出免训练框架 Stroke2Sketch,通过跨图像笔触注意力、自适应对比度增强和语义聚焦注意力,实现参考笔触属性的精确迁移与内容结构保持。

摘要:

生成由参考风格引导的草图,需要在保持语义结构和内容保真度的同时,精确迁移线条粗细、变形和纹理稀疏性等笔触属性。为此,我们提出 Stroke2Sketch,这是一种新颖的免训练框架,引入跨图像笔触注意力;该机制嵌入自注意力层中,用于建立细粒度语义对应并实现准确的笔触属性迁移。这使我们的方法能够将参考笔触特征自适应地融合到内容图像中,同时保持结构完整性。此外,我们开发了自适应对比度增强和语义聚焦注意力,以强化内容保持和前景强调。Stroke2Sketch 能有效合成风格忠实、接近手绘结果的草图,并在富有表现力的笔触控制和语义连贯性方面优于现有方法。代码位于 https://github.com/rane7/Stroke2Sketch。

英文摘要:

Generating sketches guided by reference styles requires precise transfer of stroke attributes, such as line thickness, deformation, and texture sparsity, while preserving semantic structure and content fidelity. To this end, we propose Stroke2Sketch, a novel training-free framework that introduces cross-image stroke attention, a mechanism embedded within self-attention layers to establish fine-grained semantic correspondences and enable accurate stroke attribute transfer. This allows our method to adaptively integrate reference stroke characteristics into content images while maintaining structural integrity. Additionally, we develop adaptive contrast enhancement and semantic-focused attention to reinforce content preservation and foreground emphasis. Stroke2Sketch effectively synthesizes stylistically faithful sketches that closely resemble handcrafted results, outperforming existing methods in expressive stroke control and semantic coherence. Codes are available at https://github.com/rane7/Stroke2Sketch.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑