计算机科学 ›› 2023, Vol. 50 ›› Issue (7): 129-136.doi: 10.11896/jsjkx.220700008
颜明强, 余鹏飞, 李海燕, 李红松
YAN Mingqiang, YU Pengfei, LI Haiyan, LI Hongsong
摘要： 图像风格迁移的目标是通过将目标图像风格迁移到给定的内容图像来合成输出图像。目前已有大量关于图像风格迁移的工作,但这些方法的风格化结果忽略了内容图像不同语义区域的流形分布,同时,大多数方法使用全局统计数据(如Gram矩阵或协方差矩阵)来实现风格特征到内容特征的匹配,不可避免地存在内容丢失、风格泄漏和伪影的问题,从而产生不一致的风格化结果。针对以上问题,提出了一个基于自注意力机制的渐进式流形特征映射模块(MFMM-AM),用于协调一致地匹配相关内容和风格流形之间的特征;然后通过在图像特征空间中应用精确直方图匹配(EHM)来实现风格和内容特征图的高阶分布匹配,减少了图像信息的丢失;最后,引入了两个对比性损失,利用大规模风格数据集的外部信息来学习人类感知的风格信息,使风格化图像的色彩分布和纹理图案更加合理。实验结果表明,与现有典型的任意图像风格迁移方法相比,所提网络极大地弥合了人类创作的艺术品和人工智能创作的艺术品之间的鸿沟,可以生成视觉上更加和谐和令人满意的艺术图像。
中图分类号:
| [1]LI W S,ZHAO P,YIN L Z,et al.Regional diversified image style transfer method based on Gaussian sampling [J].Journal of Computer-Aided Design & Computer Graphics,2022,34(5):8. [2]GATYS L A,ECKER A S,BETHGE M.Image style transferusing convolutional neural networks[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2016:2414-2423. [3]YIN W,YIN H,BARAKA K,et al.Dance style transfer with cross-modal transformer[C]//Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision.2023:5058-5067. [4]LI Y,FANG C,YANG J,et al.Universal style transfer via feature transforms[C]//Proceedings of the 31st International Conference on Neural Information Processing Systems.2017:385-395. [5]LI X,LIU S,KAUTZ J,et al.Learning linear transformations for fast image and video style transfer[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2019:3809-3817. [6]SAMUTH B,TSCHUMPERLÉ D,RABIN J.A Patch-BasedApproach for Artistic Style Transfer via Constrained Multi-Scale Image Matching[C]//2022 IEEE International Confe-rence on Image Processing(ICIP).IEEE,2022:3490-3494. [7]SHENG L,LIN Z,SHAO J,et al.Avatar-net:Multi-scale zero-shot style transfer by feature decoration[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2018:8242-8250. [8]KıNLı F,ÖZCAN B,KıRAÇ F.Patch-wise contrastive stylelearning for instagram filter removal[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2022:578-588. [9]LIAO J,YAO Y,YUAN L,et al.Visual attribute transferthrough deep image analogy[J].arXiv:1705.01088,2017. [10]GU S,CHEN C,LIAO J,et al.Arbitrary style transfer withdeep feature reshuffle[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2018:8222-8231. [11]KIM H,CHOI Y,KIM J,et al.Exploiting spatial dimensions of latent in gan for real-time image editing[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2021:852-861. [12]SANAKOYEU A,KOTOVENKO D,LANG S,et al.A style-aware content loss for real-time hd style transfer[C]//Procee-dings of the European Conference on Computer Vision(ECCV).2018:698-714. [13]ZHENG X A,MICHAEL WILBER B,CHEN F C,et al.Adversarial training for fast arbitrary style transfer[J].Computers & Graphics,2020,87:1-11. [14]ZHANG Y,LI M,LI R,et al.Exact feature distribution ma-tching for arbitrary style transfer and domain generalization[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2022:8035-8045. [15]AN J,HUANG S,SONG Y,et al.Artflow:Unbiased imagestyle transfer via reversible neural flows[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2021:862-871. [16]LIU S,LIN T,HE D,et al.Adaattn:Revisit attention mechanism in arbitrary neural style transfer[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision.2021:6649-6658. [17]WU Z,ZHU Z,DU J,et al.CCPL:Contrastive Coherence Preserving Loss for Versatile Style Transfer[J].arXiv:2207.04808,2022. [18]ZHANG Y,TANG F,DONG W,et al.Domain Enhanced Arbitrary Image Style Transfer via Contrastive Learning[J].arXiv:2205.09542,2022. [19]HUO J,JIN S,LI W,et al.Manifold alignment for semantically aligned style transfer [C]//Proceedings of the IEEE/CVF International Conference on Computer Vision.2021:14861-14869. [20]DING X,ZHANG X,MA N,et al.Repvgg:Making vgg-styleconvnets great again[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2021:13733-13742. [21]LI P,ZHAO L,XU D,et al.Optimal transport of deep feature for image style transfer[C]//Proceedings of the 2019 4th International Conference on Multimedia Systems and Signal Proces-sing.2019:167-171. [22]ZHANG Y,LI M,LI R,et al.Exact feature distribution ma-tching for arbitrary style transfer and domain generalization[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2022:8035-8045. [23]CHEN H,WANG Z,ZHANG H,et al.Artistic Style Transfer with Internal-external Learning and Contrastive Learning[C]//NeurIPS.2021. [24]KAMMOUN A,SLAMA R,TABIA H,et al.Generative Adversarial Networks for face generation:A survey[J].ACM Computing Surveys,2022,55(5):1-37. [25]PARK D Y,LEE K H.Arbitrary style transfer with style-attentional networks[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2019:5880-5888. [26]LIN T Y,MAIRE M,BELONGIE S,et al.Microsoft coco:Common objects in context[C]//European Conference on Computer Vision.2014:740-755. [27]PHILLIPS F,MACKINTOSH B.Wiki Art Gallery,Inc.:A case for critical thinking[J].Issues in Accounting Education,2011,26(3):593-608. [28]WANG T C,LIU M Y,ZHU J Y,et al.High-resolution image synthesis and semantic manipulation with conditional gans[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2018:8798-8807. [29]LU L H.Simulation physics-informed deep neural network byadaptive Adam optimization method to perform a comparative study of the system[J].Engineering with Computers,2022,38(Suppl 2):1111-1130. |
|
||
首页