| 摘 要: 针对当前的病毒软件检测方法难以应对大数据时代下病毒软件快速分类问题，提出一种病毒可视化检测的分类方法。详细阐述了病毒软件可视化过程，并提出一种卷积神经网络结合注意力机制的模型(即CNN_CBAM模型)进行病毒软件家族分类的深度学习方法。病毒软件样本采用BIG2015和Malimg数据集，将其进行可视化，并将CNN_CBAM模型在可视化后的数据集上进行训练。实验结果显示，CNN_CBAM模型能够有效地对病毒软件家族进行分类，且效果优于其他深度学习模型，其准确率比CNN_SVM病毒分析的方法提升16.77%。 |
| 关键词: 病毒软件；深度学习；灰度图；可视化；注意力机制 |
|
中图分类号: TP391.41
文献标识码: A
|
| 基金项目: 贵州省省级重点学科“计算机科学与技术”(ZDXK[2018]007号)；贵州省科学技术基金计划(黔科合基础[2016]1116)；贵州省教育厅创新群体研究项目(黔教合KY字[2021]022)；贵州省省级重点支持学科“计算机应用技术”(黔学位合字ZDXK[2016]20号)；贵州省2018年第三批省级服务业发展引导资金项目(黔发改服务[2018]1181号). |
|
| Visual Detection Method of Virus Software based on Attention Mechanism |
|
ZHAO Chenjie1,2, ZUO Yu1,2, CUI Zhongwei2, LI Liangliang1, WU Lian2, WANG Yongjin1,2, WEI Pingping2
|
( 1.College of Computer Science and Technology, Guizhou University, Guiyang 550025, China;
2.College of Mathematics and Big Data, Guizhou Education University, Guiyang 550018, China)
495752340@qq.com; bjuzuoyu@163.com; 33374225@qq.com; besruiven@gmail.com; 373201377@qq.com; 1411356083@qq.com; 104018794@qq.com
|
| Abstract: Current virus software detection methods have difficulty in grappling with the rapid classification of virus software in big data era. In view of this issue, this paper proposes a classification method for virus visual detection, which elaborates on the visualization process of virus software. It proposes a deep learning method of convolutional neural network combined with attention mechanism model (ie CNN_CBAM model, Convolutional Neural Network_Convolutional Block Attention Module) to classify virus software families. Virus software samples use the BIG2015 and Malimg datasets, which are visualized in this paper. The proposed CNN_CBAM model is trained on the visualized dataset. The experimental results show that the CNN_CBAM model proposed in this paper can effectively classify the virus software families, and it is better than other deep learning models. Its accuracy rate is 16.77% higher than that of CNN_SVM virus analysis method. |
| Keywords: virus software; deep learning; grayscale image; visualization; attention mechanism |
