圖像去噪在計算機視覺的許多任務中發揮著重要作用，尤其是預處理步
驟。無論是企業還是科技公司，都是非常重要的技術，而良好的去噪可以
幫助模型進一步提高準確率。其中，去除水印是一項相對複雜的任務。與
盲圖去噪、去雨等任務相比，噪聲屬於固定類型，而水印的類型是靈活
的，比如公司的 logo、文字等。作為水印，大小、形狀、顏色、透明度都
是都不同，所有這些都會對模型的學習造成巨大的障礙。

現今的深度學習技術能達到不錯的效果，但不同的任務會特定設計模
型來處理，像是影像去躁任務會有處理此任務的特定模型, 浮水印去除的
任務也有專門的模型應用到這個任務上。但由於任務的躁聲類型不同，這
兩個任務的模型彼此並不兼容。本研究專注在能同時應用影像去躁任務和
浮水印去除任務的深度學習模型。

本篇論文利用對抗性自編碼器同時處理影像去躁和浮水印去除這兩個
任務。此外有鑑於資料擴增能幫助模型學習到更多有用的特徵資訊，我們
也加入了此方法提升模型去除躁聲的能力。

Image denoising plays an important role in many tasks in computer vision,
especially the preprocessing step. Whether it is an enterprise or a technology
company, it is a very important technology, and good denoising can
help the model to further improve the accuracy. Among them, removing
the watermark is a relatively complicated task. Compared with tasks such
as blind image denoising and rain removal, the noise is of a fixed type,
while the type of watermark is flexible, such as the company’s logo, text,
etc. As watermarks, the size, shape, color, transparency are all different,
all of which can cause huge obstacles to the learning of the model.

Today’s deep learning techniques can achieve good results, but different
tasks will have different models to handle. For example, the image
denoising task will have a specific model for this task, and the watermark
removal task will also have a special model applied to this task. However,
due to the different types of noise in the tasks, the models of these two
tasks are not mutually compatible with each other. We focuses on deep
learning models that can simultaneously apply image denoising tasks and
watermark removal tasks.

This paper used an adversarial autoencoder to simultaneously handle
the two tasks of image denoising and watermark removal. In addition, since
data augmentation can help the model learn more useful feature information,
we also added this method to improve the model’s ability to remove
noise.

[1] A. Buades, B. Coll, and J.-M. Morel, “A non-local algorithm for image denoising,” in 2005 IEEE
computer society conference on computer vision and pattern recognition (CVPR’05), vol. 2, pp. 60–
65, Ieee, 2005.
[2] K. Dabov, A. Foi, V. Katkovnik, and K. Egiazarian, “Image denoising by sparse 3-d transform-domain
collaborative filtering,” IEEE Transactions on image processing, vol. 16, no. 8, pp. 2080–2095, 2007.
[3] V. Jain and S. Seung, “Natural image denoising with convolutional networks,” Advances in neural
information processing systems, vol. 21, 2008.
[4] H. C. Burger, C. J. Schuler, and S. Harmeling, “Image denoising: Can plain neural networks compete
with bm3d?,” in 2012 IEEE conference on computer vision and pattern recognition, pp. 2392–2399,
IEEE, 2012.
[5] C. Tian, Y. Xu, and W. Zuo, “Image denoising using deep cnn with batch renormalization,” Neural
Networks, vol. 121, pp. 461–473, 2020.
[6] K. Zhang, W. Zuo, Y. Chen, D. Meng, and L. Zhang, “Beyond a gaussian denoiser: Residual learning
of deep cnn for image denoising,” IEEE transactions on image processing, vol. 26, no. 7, pp. 3142–
3155, 2017.
[7] L. Wang, S. Xu, R. Xu, X. Wang, and Q. Zhu, “Non-transferable learning: A new approach for model
ownership verification and applicability authorization,” in International Conference on Learning Representations,
2021.
[8] D. Cheng, X. Li, W.-H. Li, C. Lu, F. Li, H. Zhao, and W.-S. Zheng, “Large-scale visible watermark detection
and removal with deep convolutional networks,” in Chinese conference on pattern recognition
and computer vision (prcv), pp. 27–40, Springer, 2018.
[9] X. Li, C. Lu, D. Cheng, W.-H. Li, M. Cao, B. Liu, J. Ma, and W.-S. Zheng, “Towards photo-realistic
visible watermark removal with conditional generative adversarial networks,” in International Conference
on Image and Graphics, pp. 345–356, Springer, 2019.
[10] Y. Liu, Z. Zhu, and X. Bai, “Wdnet: Watermark-decomposition network for visible watermark removal,”
in Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision,
pp. 3685–3693, 2021.
[11] J. Han, W. Li, P. Fang, C. Sun, J. Hong, M. A. Armin, L. Petersson, and H. Li, “Blind image decomposition,”
arXiv preprint arXiv:2108.11364, 2021.
[12] M. Sharma, A. Verma, and L. Vig, “Learning to clean: A gan perspective,” in Asian Conference on
Computer Vision, pp. 174–185, Springer, 2018.
[13] I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, and
Y. Bengio, “Generative adversarial nets,” Advances in neural information processing systems, vol. 27,
2014.
[14] O. Ronneberger, P. Fischer, and T. Brox, “U-net: Convolutional networks for biomedical image segmentation,”
in International Conference on Medical image computing and computer-assisted intervention,
pp. 234–241, Springer, 2015.
[15] J. Gurrola-Ramos, O. Dalmau, and T. E. Alarcón, “A residual dense u-net neural network for image
denoising,” IEEE Access, vol. 9, pp. 31742–31754, 2021.
[16] F. Jia, W. H. Wong, and T. Zeng, “Ddunet: Dense dense u-net with applications in image denoising,”
in Proceedings of the IEEE/CVF International Conference on Computer Vision, pp. 354–364, 2021.
[17] J.-Y. Zhu, T. Park, P. Isola, and A. A. Efros, “Unpaired image-to-image translation using cycleconsistent
adversarial networks,” in Proceedings of the IEEE international conference on computer
vision, pp. 2223–2232, 2017.
[18] Z. Li, J. Huang, L. Yu, Y. Chi, and M. Jin, “Low-dose ct image denoising using cycle-consistent
adversarial networks,” in 2019 IEEE Nuclear Science Symposium and Medical Imaging Conference
(NSS/MIC), pp. 1–3, IEEE, 2019.
[19] G. W. Braudaway, “Protecting publicly-available images with an invisible watermark,” in Proc. IEEE
Int. Conf. Image Processing, 1997.
[20] Z. Cao, S. Niu, J. Zhang, and X. Wang, “Generative adversarial networks model for visible watermark
removal,” IET Image Processing, vol. 13, no. 10, pp. 1783–1789, 2019.
[21] P. Isola, J.-Y. Zhu, T. Zhou, and A. A. Efros, “Image-to-image translation with conditional adversarial
networks,” in Proceedings of the IEEE conference on computer vision and pattern recognition,
pp. 1125–1134, 2017.
[22] A. Radford, L. Metz, and S. Chintala, “Unsupervised representation learning with deep convolutional
generative adversarial networks,” arXiv preprint arXiv:1511.06434, 2015.
[23] D. P. Kingma and M. Welling, “Auto-encoding variational bayes,” arXiv preprint arXiv:1312.6114,
2013.
[24] S. Ioffe, “Batch renormalization: Towards reducing minibatch dependence in batch-normalized models,”
Advances in neural information processing systems, vol. 30, 2017.
[25] Z. Liu, H. Mao, C.-Y. Wu, C. Feichtenhofer, T. Darrell, and S. Xie, “A convnet for the 2020s,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 11976–
11986, 2022.
[26] L. Deng, “The mnist database of handwritten digit images for machine learning research [best of the
web],” IEEE Signal Processing Magazine, vol. 29, no. 6, pp. 141–142, 2012.
[27] J. Wang, X. Li, and J. Yang, “Stacked conditional generative adversarial networks for jointly learning
shadow detection and shadow removal,” in Proceedings of the IEEE Conference on Computer Vision
and Pattern Recognition, pp. 1788–1797, 2018.
[28] D. P. Kingma and J. Ba, “Adam: A method for stochastic optimization,” arXiv preprint
arXiv:1412.6980, 2014.
[29] E. Schonfeld, B. Schiele, and A. Khoreva, “A u-net based discriminator for generative adversarial
networks,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition,
pp. 8207–8216, 2020.