The Key Laboratory of Advanced Design and Intelligent Computing, School of Software Engineering, Dalian University, Dalian 116622, China.
Dalian University Experimental Center, Dalian University, Dalian 116622, China.
Comput Math Methods Med. 2022 Jul 14;2022:9880038. doi: 10.1155/2022/9880038. eCollection 2022.
Deep blind watermarking algorithms based on an end-to-end encoder-decoder architecture have recently been extensively studied as an important technology for protecting copyright. However, none of the existing algorithms can fully utilize the channel features of the image to improve the robustness against JPEG compression while obtaining high visual quality. Therefore, we propose firstly a mixed-frequency channel attention method in the encoder, which utilizes different frequency components of the 2D-DCT domain as weight coefficients during channel squeezing and excitation. Its essence is to suppress the useless feature maps and enhance the feature maps suitable for watermarking embedding by introducing frequency analysis in the channel dimension. The experimental results indicate that the PSNR of our method reaches over 38 and the BER is less than 0.01% under the JPEG compression with quality factor = 50. Besides, the proposed framework also obtains excellent robustness for a variety of common distortions, including Gaussian filter, crop, crop out, and drop out.
基于端到端编解码器架构的深度盲水印算法最近作为保护版权的一项重要技术得到了广泛研究。然而，现有的算法都不能充分利用图像的信道特征，在提高对 JPEG 压缩的鲁棒性的同时获得高视觉质量。因此，我们首先在编码器中提出了一种混合频率信道注意力方法，该方法在信道压缩和激励过程中使用二维 DCT 域的不同频率分量作为权系数。其本质是通过在信道维度上进行频率分析，抑制无用的特征图，并增强适合水印嵌入的特征图。实验结果表明，在质量因子为 50 的 JPEG 压缩下，我们的方法的 PSNR 超过 38，BER 小于 0.01%。此外，所提出的框架对于包括高斯滤波、裁剪、裁剪掉和丢失在内的各种常见失真也具有出色的鲁棒性。
