| 引用本文： |
-
毛月恒,喻民,张彭博,姜建国,刘明奇,李潇萌,黄伟庆.数字多媒体应用取证技术综述[J].信息安全学报,已采用 [点击复制]
- MaoYueheng,Yumin,ZhangPengbo,JiangJianguo,LiuMingqi,LiXiaomeng,HuangWeiqing.A Survey of Digital Multimedia Application Forensics Technology[J].Journal of Cyber Security,Accept [点击复制]
|
|
| 摘要: |
| 随着互联网的广泛发展与普及，涉及数字多媒体应用的网络犯罪越来越多。多媒体应用的飞速发展在给人们带来极大便利的同时，也给电子数据取证工作带来了前所未有的挑战。有别于传统终端应用的取证，多媒体应用取证主要从多个角度，包括数字文本、数字图像、数字音频和数字视频等方面进行取证。多媒体应用取证不仅是对于媒介内容的真实性进行识别与分析，更是对媒介文件是否被篡改和伪造进行鉴别。随着人工智能的飞速发展，虚假与现实之间的差距越来越小，这对多媒体应用的取证研究带来了巨大的挑战。本文旨在对多媒体应用的优秀取证方法进行总结，详细分析了当前多媒体应用中的取证研究，首先对电子数据取证的相关概念和理论进行了介绍；其次当前多媒体应用的取证对象进行了梳理和归纳；然后将数字多媒体应用取证方法分为四类，分别是传统取证方法、机器学习取证方法、深度学习取证方法和其他取证方法，针对不同取证方法的研究进行了归类梳理与总结，并且汇总了各类取证技术的评价方法、数据集以及工具平台；最后，对多媒体应用取证未来的技术研究发展方向进行了探讨与展望。 |
| 关键词: 电子数据取证 多媒体取证 机器学习 深度学习 |
| DOI： |
| 投稿时间：2024-09-29修订日期：2025-03-14 |
| 基金项目:中国科学院青年创新促进会 |
|
| A Survey of Digital Multimedia Application Forensics Technology |
|
MaoYueheng, Yumin, ZhangPengbo, JiangJianguo, LiuMingqi, LiXiaomeng, HuangWeiqing
|
| (Institute of Information Engineering，Chinese Academy of Sciences) |
| Abstract: |
| With the extensive development and popularization of the Internet, there are more and more cybercrimes involving digital multimedia applications. While the rapid development of digital multimedia applications brings great convenience to peo-ple, it also poses unprecedented challenges to digital forensics. Different from the forensics of traditional terminal applica-tions, digital multimedia application forensics mainly involves forensic investigations from the aspects of digital text, dig-ital image, digital audio, and digital video. It not only identifies and analyzes the authenticity of media content but also identifies whether media files have been tampered with or forged. With the rapid development of artificial intelligence, the gap between fiction and reality is narrowing, posing significant challenges to the forensics research of multimedia applica-tions. This article aims to summarize the excellent forensic methods for multimedia applications and provide a detailed analysis of current research in this field. Firstly, it introduces relevant concepts and theories related to digital forensics. Secondly, it organizes and summarizes the forensic objects of current digital multimedia applications. Then, digital multi-media application forensics methods are divided into four categories, traditional forensics methods, machine learning fo-rensics methods, deep learning forensics methods and other forensics methods. The research on different forensics meth-ods is classified and summarized, and the evaluation methods, data sets and tool platforms of various forensics technolo-gies are summarized. Finally, it discusses and forecasts the future research and development directions of multimedia ap-plication forensics. |
| Key words: digital forensics digital multimedia forensics machine learning deep learning |
