本文是一份关于风格迁移的调研，涵盖了从最初的开创性工作到现在的多项技术，包括使用神经网络实现的艺术风格转换，实例归一化在快速风格转换中的应用，以及实时任意风格转移等。文章列举了相关论文、实践案例、阅读笔记和代码实现，提供了深入理解和实现风格迁移的资源。
style_transfer_survey
A survey on style_transfer from the original fantasy paper till now.
Contents:
Papers
- A Neural Algorithm of Artistic Style
- arxiv: 1508.06576
- github: https://github.com/jcjohnson/neural-style
- translation: https://www.jianshu.com/p/9f03b61fdeac
- Texture Networks: Feed-forward Synthesis of Textures and Stylized Images
- arxiv: 1603.03417, also ICML 2016
- github: author torch,
- translation: https://www.jianshu.com/p/1187049ae1ad
- Perceptual Losses for Real-Time Style Transfer and Super-Resolution
- arxiv: 1603.08155, also ECCV 2016
- github: author torch, third pytorch
- translation: https://www.jianshu.com/p/b728752a70e9
- webpage: project page
- Incorporating long-range consistency in CNN-based texture generation
- arxiv: 1606.01286
- Instance Normalization: The missing Ingredient for Fast Stylization
- arxiv:
最低0.47元/天 解锁文章
扫一扫
799
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
