Journal of Computer Applications ›› 2024, Vol. 44 ›› Issue (7): 2102-2109.DOI: 10.11772/j.issn.1001-9081.2023070919

• Cyber security • Previous Articles     Next Articles

Generative data hiding algorithm based on multi-scale attention

Li LIU(), Haijin HOU, Anhong WANG, Tao ZHANG   

  1. School of Electronic Information and Engineering,Taiyuan University of Science and Technology,Taiyuan Shanxi 030024,China
  • Received:2023-07-11 Revised:2023-09-14 Accepted:2023-09-19 Online:2023-10-26 Published:2024-07-10
  • Contact: Li LIU
  • About author:HOU Haijin, born in 1997. M. S. candidate. His research interests include information hiding.
    WANG Anhong, born in 1972. Ph. D., professor. Her research interests include image/video coding, information security.
    ZHANG Tao, born in 1999. M. S. candidate. His research interests include deep learning, information hiding.
    First author contact:LIU Li, born in 1978. Ph. D., associate professor. Her research interests include information hiding.
  • Supported by:
    National Natural Science Foundation of China(62072325);Fundamental Research Program of Shanxi Province(202103021224272);Scientific Research Initiation Fund of Taiyuan University of Science and Technology(20212039)

基于多尺度注意力的生成式信息隐藏算法

刘丽(), 侯海金, 王安红, 张涛   

  1. 太原科技大学 电子信息工程学院,太原 030024
  • 通讯作者: 刘丽
  • 作者简介:侯海金(1997—),男,山西晋中人,硕士研究生,主要研究方向:信息隐藏;
    王安红(1972—),女,山西运城人,教授,博士,CCF会员,主要研究方向:图像/视频编码、信息安全;
    张涛(1999—),男,山西吕梁人,硕士研究生,主要研究方向:深度学习、信息隐藏。
    第一联系人:刘丽(1978—),女,宁夏吴忠人,副教授,博士,主要研究方向:信息隐藏;
  • 基金资助:
    国家自然科学基金资助项目(62072325);山西省基础研究计划项目(202103021224272);太原科技大学科研启动基金资助项目(20212039)

Abstract:

Aiming to the problems of low embedding capacity and poor visual quality of the extracted secret images in existing generative data hiding algorithms, a generative data hiding algorithm based on multi-scale attention was proposed. First, a generator with dual encode-single decode based on multi-scale attention was designed. The features of the cover image and secret image were extracted independently at the encoding end in two branches, and fused at the decoding end by a multi-scale attention module. Skip connections were used to provide different scales of detail features, thereby ensuring high-quality of the stego-image. Second, self-attention module was introduced into the extractor of the U-Net structure to weaken the deep features of the cover image and enhance the deep features of the secret image. The skip connections were used to compensate for the detail features of the secret image, so as to improve the accuracy of the extracted secret data. At the same time, the adversarial training of the multi-scale discriminator and generator could effectively improve the visual quality of the stego-image. Experimental results show that the proposed algorithm can achieve an average Peak Signal-to-Noise Ratio (PSNR) and Structure Similarity Index Measure (SSIM) of 40.93 dB and 0.988 3 for the generated stego-images, and an average PSNR and SSIM of 30.47 dB and 0.954 3 for the extracted secret images under the embedding capacity of 24 bpp.

Key words: data hiding, attention mechanism, multi-scale, encode-decode structure, Generative Adversarial Network (GAN)

摘要:

针对现有生成式信息隐藏算法嵌入容量低且提取的秘密图像视觉质量欠佳的问题,提出基于多尺度注意力的生成式信息隐藏算法。首先,设计基于多尺度注意力的双编码-单解码生成器,载体图像与秘密图像的特征在编码端分两个支路独立提取,在解码端通过多尺度注意力模块进行融合,并利用跳跃连接为解码端提供不同尺度的细节特征,从而获得高质量的载密图像。其次,在U-Net结构的提取器中引入自注意力模块,以弱化载体图像特征、增强秘密图像深层特征,并利用跳跃连接弥补秘密图像细节特征,提高秘密信息提取的准确率;同时,多尺度判决器与生成器的对抗训练可以有效提升载密图像的视觉质量。实验结果表明,所提算法在嵌入容量为24 bpp的情况下,生成的载密图像峰值信噪比(PSNR)和结构相似性(SSIM)平均可达到40.93 dB和0.988 3,且提取的秘密图像PSNR和SSIM平均可达到30.47 dB和0.954 3。

关键词: 信息隐藏, 注意力机制, 多尺度, 编码-解码结构, 生成对抗网络

CLC Number: