摘要
隐写术是实现隐蔽通信的一种手段,而隐写分析是用来检测是否存在秘密信息隐蔽传输的技术,两者在相互对抗中不断进步与发展。基于失真函数和校验网格码(STC)结合的自适应图像隐写算法的提出使得图像隐写分析愈加困难,导致隐写分析算法难以对图像隐写区域进行针对性检测。为此,许多专家提出在基于深度学习的隐写分析模型中加入注意力机制,引导模型重点关注隐写区域的特征,从而实现检测准确率提升的目标。本文介绍了近几年在隐写分析模型中引入各种注意力机制来提高模型性能的技术,对最新的技术进行了剖析,总结和展望该机制在隐写分析中的研究前景和应用方向,为后续研究提供有价值的参考依据。
Steganography is a means to achieve covert communication,and steganalysis is a technology used to detect whether secret information is covertly transmitted.Both of them continue to progress and develop in the confrontation with each other.The proposal of an adaptive image steganography algorithm based on the combination of distortion function and Syndrome-Trellis Code(STC)makes image steganalysis increasingly difficult,and steganalysis algorithms find it difficult to perform targeted detection of stego regions.So many experts propose to add attention mechanism to the steganalysis model based on deep learning to guide the model to pay more attention to the characteristics of the stego region,so as to improve the detection accuracy.This paper introduces the technology of adding various attention mechanisms to steganalysis model to improve the performance of the model in recent years,analyzes the latest technology,summarizes and forecasts the research prospect and application direction of this mechanism in steganalysis,and provides a valuable reference for the following research.
作者
吴朝平
杨本娟
Wu Chaoping;Yang Benjuan(School of Mathematical Sciences,Guizhou Normal University,Guiyang,Guizhou 550025,China)
出处
《计算机时代》
2026年第3期27-33,共7页
Computer Era
基金
2025年度贵州省高等学校本科教学内容和课程体系改革项目(GZJG2025083)。
