 |
|
|
|
 |
 |
 |
|
 |
|
 |
|
|
李旭嵘,纪守领,吴春明,刘振广,邓水光,程鹏,杨珉,孔祥维.深度伪造与检测技术综述.软件学报,2021,32(2):0 |
深度伪造与检测技术综述 |
Survey on Deepfakes and Detection Techniques |
投稿时间:2020-05-07 修订日期:2020-06-22 |
DOI:10.13328/j.cnki.jos.006140 |
中文关键词: 深度学习 深度伪造 假视频 取证 检测技术 |
英文关键词:deep learning deepfakes fake videos forensics detection techniques |
基金项目:国家重点研发计划项目(2018YFB0804102,2020YFB1804705);浙江省自然科学基金杰出青年项目(LR19F020003);浙江省重点研发计划项目(2019C01055,2020C01021);国家自然科学基金项目(61772466,U1936215,U1836202),前沿科技创新专项(2019QY(Y)0205) |
作者 | 单位 | E-mail | 李旭嵘 | 浙江大学计算机科学与技术学院, 杭州 310027 阿里巴巴, 杭州 310023 | | 纪守领 | 浙江大学计算机科学与技术学院, 杭州 310027 | sji@zju.edu.cn | 吴春明 | 浙江大学计算机科学与技术学院, 杭州 310027 之江实验室, 杭州 310000 | | 刘振广 | 浙江工商大学计算机与信息工程学院, 杭州 310027 | | 邓水光 | 浙江大学计算机科学与技术学院, 杭州 310027 | | 程鹏 | 浙江大学控制科学与工程学院, 杭州 310027 | | 杨珉 | 复旦大学计算机科学与技术学院, 上海 201203 | | 孔祥维 | 浙江大学管理学院, 杭州 310027 | |
|
摘要点击次数: 1203 |
全文下载次数: 1955 |
中文摘要: |
深度学习在计算机视觉领域取得了重大成功,超越了众多传统的方法.然而,近年来深度学习技术被滥用在假视频的制作上,使得以Deepfakes为代表的伪造视频在网络上泛滥成灾.这种深度伪造技术通过篡改或替换原始视频的人脸信息,并合成虚假的语音,来制作色情电影、虚假新闻、政治谣言等.为了消除此类伪造技术带来的负面影响,众多学者对假视频的鉴别进行了深入的研究,并提出一系列的检测方法帮助机构或社区来识别此类伪造视频.尽管如此,目前的检测技术仍然存在依赖特定分布数据、特定压缩率等众多的局限性,远远落后于假视频的生成技术.并且,不同的学者解决问题的角度不同,使用的数据集和评价指标均不统一.迄今为止,学术界对深度伪造与检测技术仍缺乏统一的认识,深度伪造和检测技术研究的体系架构尚不明确.在本综述中,我们回顾了深度伪造与检测技术的发展,并对现有研究工作进行了系统的总结和科学的归类.最后,我们讨论了深度伪造技术蔓延带来的社会风险,分析了检测技术的诸多局限性,并探讨了检测技术面临的挑战和潜在研究方向,旨在为后续学者进一步推动深度伪造检测技术的发展和部署提供指导. |
英文摘要: |
Deep learning has achieved great success in in the field of computer vision, surpassing many traditional methods. However, in recent years, deep learning technology has been abused in the production of fake videos, making fake videos represented by Deepfakes flooding on the Internet. This technique produces pornographic movies, fake news, political rumors by tampering or replacing the face information of the original videos and synthesize fake speech. In order to eliminate the negative effects brought by such forgery technologies, many researchers have conducted in-depth research on the identification of fake videos and proposed a series of detection methods to help institutions or communities to identify such fake videos. Nevertheless, the current detection technology still has many limitations such as specific distribution data, specific compression ratio, and so on, far behind the generation technology of fake video. In addition, different researchers handle the problem from different angles. The data sets and evaluation indicators used are not uniform. So far, the academic community still lacks a unified understanding of deep forgery and detection technology. The architecture of deep forgery and detection technology research is not clear. In this review, we review the development of deep forgery and detection technologies. Besides, we systematically summarize and scientifically classify existing research works. Finally, we discussed the social risks posed by the spread of Deepfakes technology, analyzed the limitations of detection technology, and discussed the challenges and potential research directions of detection technology, aiming to provide guidance for follow-up researchers to further promote the development and deployment of Deepfakes detection technology. |
HTML 下载PDF全文 查看/发表评论 下载PDF阅读器 |
|
|
|
|
|
|
 |
|
|
|
|
 |
|
 |
|
 |
|