多尺度语义信息无监督山水画风格迁移网络

doi:10.3778/j.issn.1002-8331.2210-0079

摘要/Abstract

摘要： 针对图像转换类的生成对抗网络在处理无监督风格迁移任务时存在的纹理杂乱、生成图像质量差的问题，基于循环一致性损失提出了循环矫正多尺度评估生成对抗网络。首先在网络架构的设计上，基于图像的三层语义信息提出了多尺度评估网络架构，以此强化源域到目标域的迁移效果；其次在损失函数的改进上，提出了多尺度对抗损失以及循环矫正损失，用于以更严苛的目标引导模型的迭代优化方向，生成视觉质量更好的图片；最后为了预防模式崩溃的问题，在风格特征的编码阶段添加了注意力机制以提取重要的特征信息，在网络的各阶段引入ACON激活函数以加强网络的非线性表达能力，避免神经元坏死。实验结果表明，相比于CycleGAN、ACL-GAN，所提出方法在山水画风格迁移数据集上的FID值分别降低了21.80%和34.33%；为了验证模型的泛化能力，在Vangogh2Photo和Monet2Photo两个公开数据集上进行了泛化实验对比，FID值相比于两个对照网络分别降低了7.58%、18.14%和4.65%、6.99%。

关键词: 无监督风格迁移, 生成对抗网络（GAN）, 多尺度评估, CycleGAN

Abstract: This paper proposes CCME-GAN (circulatory correction multiscale evaluation-generative adversarial networks) based on the cycle consistency loss, aiming at the problems of texture clutter and poor quality of generated images when the generative adversarial network of image conversion class is dealing with the task of unsupervised style transfer. Firstly, in the design of the network architecture, a multiscale evaluation network architecture based on the three-layer semantic information of images is proposed to enhance the transfer effect from the source domain to the target domain. Secondly, in the improvement of the loss function, a multiscale adversarial loss and a cyclic correction loss are proposed to guide the optimization iteration direction of the model with a stricter target, and generate pictures with better visual quality. Finally, in order to prevent the problem of pattern collapse, this paper adds an attention mechanism in the encoding stage of style features to extract important feature information, and then introduces the ACON activation function in each stage of the network to strengthen the nonlinear expression ability of the network and avoid neuron necrosis. The experimental results show that the FID value of this paper method is reduced by 21.80% and 34.33% compared with CycleGAN and ACL-GAN on the landscape painting style migration dataset. In addition, in order to verify the generalization ability of the model, the generalization experiments are compared on two public datasets, Vangogh2Photo, and Monet2Photo and the FID values are decreased by 7.58%, 18.14% and 4.65%, 6.99% respectively.

Key words: unsupervised style transfer, generative adversarial networks (GAN), multiscale evaluation, cycleGAN

周粤川, 张建勋, 董文鑫, 高林枫, 倪锦园. 多尺度语义信息无监督山水画风格迁移网络[J]. 计算机工程与应用, 2024, 60(4): 258-269.

ZHOU Yuechuan, ZHANG Jianxun, DONG Wenxin, GAO Linfeng, NI Jinyuan. Unsupervised Landscape Painting Style Transfer Network with Multiscale Semantic Information[J]. Computer Engineering and Applications, 2024, 60(4): 258-269.

参考文献

[1] JING Y, YANG Y, FENG Z, et al. Neural style transfer: a review[J]. IEEE Transactions on Visualization and Computer Graphics, 2019, 26(11): 3365-3385.
[2] GATYS L A, ECKER A S, BETHGE M. A neural algorithm of artistic style[J]. arXiv:1508.06576, 2015.
[3] 唐稔为, 刘启和, 谭浩. 神经风格迁移模型综述[J]. 计算机工程与应用, 2021, 57(19): 32-43.
TANG R W, LIU Q H, TAN H. A review of neural style transfer models[J]. Computer Engineering and Applications, 2021, 57(19): 32-43.
[4] JOHNSON J, ALAHI A, FEI-FEI L. Perceptual losses for real-time style transfer and super-resolution[C]//Proceedings of the 14th European Conference on Computer Vision. Cham: Springer, 2016: 694-711.
[5] ULYANOV D, LEBEDEV V, VEDALDI A, et al. Texture networks: feed-forward synthesis of textures and stylized images[C]//Proceedings of the 33rd International Conference on Machine Learning, 2016: 1349-1357.
[6] ULYNOV D, VEDALDI A, LEMPITSKY V. Improved texture networks: maximizing quality and diversity in feed-forward stylization and texture synthesis[C]//Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition, 2017: 6924-6932.
[7] DUMOULIN V, SHLENS J, KUDLUR M. A learned representation for artistic style[J]. arXiv:1610.07629, 2016.
[8] HUANG X, BELONGIE S. Arbitrary style transfer in real-time with adaptive instance normalization[C]//Proceedings of the 2017 IEEE International Conference on Computer Vision, 2017: 1501-1510.
[9] LI Y, FANG C, YANG J, et al. Universal style transfer via feature transforms[C]//Advances in Neural Information Processing Systems 30, 2017.
[10] PARK D Y, LEE K H. Arbitrary style transfer with style-attentional networks[C]//Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2019: 5880-5888.
[11] GOODFELLOW I, POUGET-ABADIE J, MIRZA M, et al. Generative adversarial nets[C]//Advances in Neural Information Processing Systems 27, 2014.
[12] MIRZA M, OSINDERO S. Conditional generative adversarial nets[J]. arXiv:1411. 1784, 2014.
[13] RADFORD A, METZ L, CHINTALA S. Unsupervised representation learning with deep convolutional generative adversarial networks[J]. arXiv: 1511. 06434, 2015.
[14] ISOLA P, ZHU J Y, ZHOU T, et al. Image-to-image translation with conditional adversarial networks[C]//Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition, 2017: 1125-1134.
[15] ZHU J Y, PARK T, ISOLA P, et al. Unpaired image-to-image translation using cycle-consistent adversarial networks[C]//Proceedings of the 2017 IEEE International Conference on Computer Vision, 2017: 2223-2232.
[16] YI Z, ZHANG H, TAN P, et al. DualGAN: unsupervised dual learning for image-to-image translation[C]//Proceedings of the 2017 IEEE International Conference on Computer Vision, 2017: 2849-2857.
[17] KIM T, CHA M, KIM H, et al. Learning to discover cross-domain relations with generative adversarial networks[C]//Proceedings of the 34th International Conference on Machine Learning, 2017: 1857-1865.
[18] ZHAO Y, WU R, DONG H. Unpaired image-to-image translation using adversarial consistency loss[C]//Proceedings of the 16th European Conference on Computer Vision. Cham: Springer, 2020: 800-815.
[19] 林锦, 陈昭炯, 叶东毅. 局部色彩可控的中国山水画仿真生成方法[J]. 小型微型计算机系统, 2021, 42(9): 1985-1991.
LIN J, CHEN Z J, YE D Y. Local color controllable Chinese landscape painting simulation generation method[J]. Journal of Chinese Computer Systems, 2021, 42(9): 1985-1991.
[20] 滕少华, 袁萧勇, 张巍. 生成对抗网络的素描生成方法[J]. 小型微型计算机系统, 2022, 43(4): 852-857.
TENG S H, YUAN X Y, ZHANG W. Sketch generation method based on generative adversarial network[J]. Journal of Chinese Computer Systems, 2022, 43(4): 852-857.
[21] WOO S, PARK J, LEE J Y, et al. CBAM: convolutional block attention module[C]//Proceedings of the 15th European Conference on Computer Vision. Cham: Springer, 2018: 3-19.
[22] MA N, ZHANG X, LIU M, et al. Activate or not: learning customized activation[C]//Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021: 8032-8042.
[23] YANG G, ZHAO H, SHI J, et al. SegStereo: exploiting semantic information for disparity estimation[C]//Proceedings of the 15th European Conference on Computer Vision. Cham: Springer, 2018: 636-651.
[24] LIU Y Q, DU X, SHEN H L, et al. Estimating generalized gaussian blur kernels for out-of-focus image deblurring[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2020, 31(3): 829-843.
[25] HU J, SHEN L, SUN G. Squeeze-and-excitation networks[C]//Proceedings of the 2018 IEEE Conference on Computer Vision and Pattern Recognition, 2018: 7132-7141.
[26] HU J, SHEN L, ALBANIE S, et al. Gather-excite: exploiting feature context in convolutional neural networks[C]//Advances in Neural Information Processing Systems 31, 2018.
[27] XUE A. End-to-end chinese landscape painting creation using generative adversarial networks[C]//Proceedings of the 2021 IEEE/CVF Winter Conference on Applications of Computer Vision, 2021: 3863-3871.