S2WAT: Image Style Transfer via Hierarchical Vision Transformer Using Strips Window Attention
Abstract
Transformer's recent integration into style transfer leverages its proficiency in establishing long-range dependencies, albeit at the expense of attenuated local modeling. This paper introduces Strips Window Attention Transformer (S2WAT), a novel hierarchical vision transformer designed for style transfer. S2WAT employs attention computation in diverse window shapes to capture both short- and long-range dependencies. The merged dependencies utilize the "Attn Merge" strategy, which adaptively determines spatial weights based on their relevance to the target. Extensive experiments on representative datasets show the proposed method's effectiveness compared to state-of-the-art (SOTA) transformer-based and other approaches. The code and pre-trained models are available at https://github.com/AlienZhang1996/S2WAT.
Cite
Text
Zhang et al. "S2WAT: Image Style Transfer via Hierarchical Vision Transformer Using Strips Window Attention." AAAI Conference on Artificial Intelligence, 2024. doi:10.1609/AAAI.V38I7.28529Markdown
[Zhang et al. "S2WAT: Image Style Transfer via Hierarchical Vision Transformer Using Strips Window Attention." AAAI Conference on Artificial Intelligence, 2024.](https://mlanthology.org/aaai/2024/zhang2024aaai-s-a/) doi:10.1609/AAAI.V38I7.28529BibTeX
@inproceedings{zhang2024aaai-s-a,
title = {{S2WAT: Image Style Transfer via Hierarchical Vision Transformer Using Strips Window Attention}},
author = {Zhang, Chiyu and Xu, Xiaogang and Wang, Lei and Dai, Zaiyan and Yang, Jun},
booktitle = {AAAI Conference on Artificial Intelligence},
year = {2024},
pages = {7024-7032},
doi = {10.1609/AAAI.V38I7.28529},
url = {https://mlanthology.org/aaai/2024/zhang2024aaai-s-a/}
}