典型文献
Double-Layer Positional Encoding Embedding Method for Cross-Platform Binary Function Similarity Detection
文献摘要：
The similarity detection between two cross-platform binary functions has been applied in many fields,such as vulnerability detection,software copyright protection or malware classification.Current advanced methods for binary function similarity detection usually use semantic features,but have certain limitations.For example,practical applications may encounter instruc-tions that have not been seen in training,which may eas-ily cause the out of vocabulary(OOV)problem.In addi-tion,the generalization of the extracted binary semantic features may be poor,resulting in a lower accuracy of the trained model in practical applications.To overcome these limitations,we propose a double-layer positional encoding based transformer model(DP-Transformer).The DP-Transformer's encoder is used to extract the semantic fea-tures of the source instruction set architecture(ISA),which is called the source ISA encoder.Then,the source ISA encoder is fine-tuned by the triplet loss while the tar-get ISA encoder is trained.This process is called DP-MIRROR.When facing the same semantic basic block,the embedding vectors of the source and target ISA en-coders are similar.Different from the traditional trans-former which uses single-layer positional encoding,the double-layer positional encoding embedding can solve the OOV problem while ensuring the separation between in-structions,so it is more suitable for the embedding of as-sembly instructions.Our comparative experiment results show that DP-MIRROR outperforms the state-of-the-art approach,MIRROR,by about 35％in terms of precision at 1.
文献关键词：
中图分类号：
作者姓名：
JIANG Xunzhi;WANG Shen;YU Xiangzhan;GONG Yuxin
作者机构：
School of Cyberspace Science,Harbin Institute of Technology,Harbin 150001,China
文献出处：
引用格式：
[1]JIANG Xunzhi;WANG Shen;YU Xiangzhan;GONG Yuxin-.Double-Layer Positional Encoding Embedding Method for Cross-Platform Binary Function Similarity Detection)[J].电子学报（英文）,2022(04):604-611
A类：
instruc,MIRROR,structions
B类：
Double,Layer,Positional,Encoding,Embedding,Method,Cross,Platform,Binary,Function,Similarity,Detection,similarity,detection,between,two,cross,platform,binary,functions,has,been,applied,many,fields,such,vulnerability,software,copyright,protection,malware,classification,Current,advanced,methods,usually,semantic,features,but,have,certain,limitations,For,example,practical,applications,may,encounter,that,not,seen,training,which,eas,ily,cause,vocabulary,OOV,problem,In,addi,generalization,extracted,poor,resulting,lower,accuracy,trained,model,To,overcome,these,propose,double,layer,positional,encoding,transformer,DP,Transformer,encoder,used,source,set,architecture,ISA,called,Then,fine,tuned,by,triplet,loss,while,This,process,When,facing,same,basic,block,embedding,vectors,target,coders,Different,from,traditional,uses,single,can,solve,ensuring,separation,more,suitable,sembly,instructions,Our,comparative,experiment,results,show,outperforms,state,art,approach,about,terms,precision
AB值：
0.552998
相似文献
机标中图分类号，由域田数据科技根据网络公开资料自动分析生成，仅供学习研究参考。
