网络与信息安全学报2024，Vol.10Issue(6)：96-108,13.DOI:10.11959/j.issn.2096-109x.2024083
基于跨流注意力增强中心差分卷积网络的CG图像检测
Cross-stream attention enhanced central difference convolutional network for CG image detection
摘要
Abstract
With the maturation of computer graphics(CG)technology in the field of image generation,the realism of created images has been improved significantly.Although these technologies are widely used in daily life and bring many conveniences,they also come with many security risks.If forged images generated using CG technol-ogy are maliciously used and widely spread on the Internet and social media,they may harm the rights of individu-als and enterprises.Therefore,an innovative cross-stream attention enhanced central difference convolutional net-work was proposed,aiming at improving the accuracy of CG image detection.A dual-stream structure was con-structed in the model,in order to extract semantic features and non-semantic residual texture features from the im-age.Vanilla convolutional layers in each stream were replaced by central difference convolutions,which allowed the model to simultaneously extract pixel intensity information and pixel gradient information from the image.Fur-thermore,by introducing a cross-stream attention enhancement module,the model enhanced feature extraction ca-pability at the global level and promoted complementarity between the two feature streams.Experimental results demonstrate that this method outperforms existing methods.Additionally,a series of ablation experiments further verify the rationality of the proposed model design.关键词
计算机图形学/CG图像检测/中心差分卷积/注意力机制Key words
computer graphics/CG image detection/central difference convolution/attention mechanism分类
信息技术与安全科学引用本文复制引用
黄锦坤,黄远航,黄文敏,骆伟祺..基于跨流注意力增强中心差分卷积网络的CG图像检测[J].网络与信息安全学报,2024,10(6):96-108,13.基金项目
国家自然科学基金(62472458) The National Natural Science Foundation of China(62472458) （62472458）
