基于生成对抗网络的服装属性控制方法
A Generative Adversarial Network-Based Approach to Garment Attribute Control
DOI: 10.12677/MOS.2023.126509, PDF, 下载: 143  浏览: 205 
作者: 晏金鋆, 向 忠, 钱 淼：浙江理工大学机械工程学院，浙江 杭州
关键词: 生成对抗网络；图像处理；图像属性控制；注意力机制；Generative Adversarial Network； Image Processing； Image Attribute Control； Attention Mechanism
摘要: 在服装设计领域中，图像生成技术帮助服装设计师能够根据市场需求快速调整款式，针对目前基于生成对抗网络的图像属性控制模型中图像语义信息难以被充分利用的问题，本文提出了一种新的基于生成对抗网络的服装图像属性控制模型，实现对服装不同属性的控制。采用了基于Unet++结构实现图像的编码与解码，相比普通的编解码器，能有效减少图像语义信息在编解码过程中的丢失，提高生成图像的质量；同时，使用CSAM注意力模块加强判别网络对输入特征通道域与空间域上的关注度，提高判别器网络对服装图像大范围属性的鉴别能力。对比了本文与其他生成模型在服装属性控制上的效果，本文所提出模型与StarGAN、AttGAN等主流图像属性控制模型相比，能够对服装图像属性进行更好地控制，并生成高质量的图像。
Abstract: In the field of clothing design, the image generation technology to help clothing designers can quickly adjust style according to market demand, in view of the current based on image semantic information which is difficult to make full use of network control model, this paper puts forward a new clothing based on the generated against network image attribute control model to realize the control of different attributes of clothing. Unet++ structure is adopted to realize image coding and decoding, compared with the ordinary codec, it can effectively reduce the loss of image semantic in-formation and improve the quality of the generated image. At the same time, the CSAM attention module is used to strengthen the discriminator network’s attention to the input feature channel domain and spatial domain, and improve the discriminator network’s ability to identify a wide range of attributes of garment images. Compared with the effect of this paper and other generative models in the clothing attribute control, compared with the mainstream image attribute control models such as StarGAN and AttGAN, the proposed model can better control the clothing image at-tributes and generate high-quality images.
文章引用：晏金鋆, 向忠, 钱淼. 基于生成对抗网络的服装属性控制方法[J]. 建模与仿真, 2023, 12(6): 5608-5620. https://doi.org/10.12677/MOS.2023.126509

参考文献

[1] 施倩, 罗戎蕾. 基于生成对抗网络的服装图像生成研究进展[J]. 现代纺织技术, 2023, 31(2): 36-46.
[2] 邓欣. 服装款式部件的数据库构建及应用[J]. 天津纺织科技, 2016(4): 32-33.
[3] 谢雪勇, 张辉. XML在服装部件信息数据存储上的优势[J]. 纺织科技进展, 2012(2): 84-86.
[4] Mirza, M. and Osindero, S. (2014) Conditional Generative Adversarial Nets.
[5] Choi, Y., Choi, M., Kim, M., et al. (2018) Stargan: Unified Generative Adversarial Networks for Multi-Domain Image-to-Image Translation. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Salt Lake City, 18-23 June 2018, 8789-8797.
https://doi.org/10.1109/CVPR.2018.00916
[6] Zhou, Z., Siddiquee, M.M.R., Tajbakhsh, N., et al. (2019) Unet++: Redesigning Skip Connections to Exploit Multiscale Fea-tures in Image Segmentation. IEEE Transactions on Medical Imaging, 39, 1856-1867.
https://doi.org/10.1109/TMI.2019.2959609
[7] Isola, P., Zhu, J.-Y., Zhou, T., et al. (2017) Image-to-Image Translation with Conditional Adversarial Networks. The Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, 21-26 July 2017, 5967-5976.
https://doi.org/10.1109/CVPR.2017.632
[8] Demir, U. and Unal, G. (2018) Patch-Based Image Inpainting with Generative Adversarial Networks.
[9] Han, X., Wu, Z., Wu, Z., et al. (2018) Viton: An Image-Based Virtual Try-On Network. The Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Salt Lake City, 18-23 June 2018, 7543-7552.
https://doi.org/10.1109/CVPR.2018.00787
[10] Simonyan, K. and Zisserman, A. (2014) Very Deep Convolutional Networks for Large-Scale Image Recognition.
[11] Heusel, M., Ramsauer, H., Unterthiner, T., et al. (2017) GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium. Proceedings of the 31st International Conference on Neural Information Processing Systems, Long Beach, 4-9 December 2017, 6629-6640.
[12] Szegedy, C., Vanhoucke, V., Ioffe, S., et al. (2016) Rethinking the Inception Architecture for Computer Vision. The Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, 27-30 June 2016, 2818-2826.
https://doi.org/10.1109/CVPR.2016.308
[13] Hu, J., Shen, L. and Sun, G. (2018) Squeeze-and-Excitation Net-works. The Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Salt Lake City, 18-23 June 2018, 7132-7141.
https://doi.org/10.1109/CVPR.2018.00745
[14] Zhang, H., Goodfellow, I., Metaxas, D., et al. (2019) Self-Attention Generative Adversarial Networks. Proceedings of the International Conference on Machine Learning, 2019, 7354-7363.
[15] Woo, S., Park, J., Lee, J.-Y., et al. (2018) CBAM: Convolutional Block Attention Module. The Proceedings of the European Conference on Computer Vision (ECCV), Munich, 8-14 September 2018, 3-19.
https://doi.org/10.1007/978-3-030-01234-2_1

为你推荐