딥러닝 기반 스테그어낼러시스와 적대적 스테가노그래피 기술의 최신 연구 동향
Recent Research Trends in Deep Learning-Based Steganalysis and Adversarial Steganography Techniques
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
최근 인공지능의 발전은 급속히 진행되고 있으며 컴퓨터 통신분야 뿐만 아니라사회 전반에 걸쳐 큰 영향을 주고 있다. 인공지능을 통해 많은 사람들은 생활에편리함을 느끼고 있으나 이를 악용하는 사례 또한 증가하고 있다. 특히, 스테가노그래피(Steganography)는 정보를 은닉하는 기술로, 해커와 간첩뿐만아니라 악성 SW 간의 은밀한 통신에 지속적으로 악용되고 있으며 기법 역시고도화되고 있다. 한편, 스테그어낼러시스(Steganalysis)는 스테가노그래피 기술로은닉된 비밀 메시지를 탐지하고 추출하는 것을 목표로 하는 기술이다. 점점 더정교해지는 은닉 기법에 대응하기 위해, 최근에는 딥러닝(Deep Learning)을 활용한스테그어낼러시스가 효과적인 대응기술로 주목받고 있다. 본 논문에서는 국방영역에서의 스테가노그래피 은닉 통신을 탐지하여 대응하기 위한 대표적인 딥러닝기반의 스테그어낼러시스 기법을 소개한 후 이를 회피하기 위한 적대적스테가노그래피 기술 동향 분석을 통해 위험성 알리고 대응 방안을 제시하고자한다.
Abstract
Recently, the development of artificial intelligence is progressing rapidly and is having a significant impact not only on the computer communication field but also on society as a whole. Many people feel convenience in their lives through artificial intelligence, but cases of abuse of it are also increasing. In particular, technologies such as steganography are continuously being abused due to their ability to hide information, and the technology and attack techniques are improving day by day. Steganalysis aims to detect and extract secret messages hidden in these messages. In particular, stagnation using deep learning has recently been attracting attention in order to respond to increasingly sophisticated hiding techniques. In this paper, we study various deep learning models such as CNN and GAN and explore methods applied to steganalysis. In addition, we will confirm the possibility of practical deep learning-based steganalysis to respond to steganography threats in defense security, and seek ways in which the technology can contribute to the defense information protection system.
참고문헌 (36)
- Wafa M. Eid, "Digital Image Steganalysis: Current Methodologies and Future Challenges", IEEE Access, 10, pp.92321~92336 (2022)
- 김민지, "CNN 기반 이미지 스테그아날리시스 모델 우회를 위한 JSMA 기반의 새로운 적대적스테고 이미지 생성 기법", 석사, pp.2024
- 육심언, "다양한 통신환경에서 스테가노그래피 기술 기반의 새로운 은닉통신 기법", 박사, pp.2023
- 육심언, "SNS 메신저 환경에서의 썸네일 이미지 기반의 새로운 스테가노그래피 통신기법 연구", Journal of Internet Computing and Services(JICS), pp.151~162 (2021)
- 박홍진, "인공지능, 기계학습, 딥 러닝 분야의 국내 논문 동향 분석", 한국정보전자통신기술학회논문지, 13(4) (2020)
- Alexandr Kuznetsov, "Image steganalysis using deep learning models", 2022 IEEE 9th International Conference on Problems of Infocommunications, Science and Technology (PIC S&T), pp.24~ (2022)
- H. Chen, "Deep learning image steganalysis method fused with CBAM", International Conference on Computer Engineering and Networks, pp.1175~1184 (2022), Springer
- Kristian D. Michaylov, "Steganography and steganalysis for digital image enhanced Forensic analysis and recommendations", Taylor & Francis Journal of Cyber Security Technology, 27 (2024)
- Y. Y. Lu, "Importance of truncation activation in pre-processing for spatial and JPEG image steganalysis", IEEE International Conference on Image Processing (ICIP), pp.689~693 (2019)
- Jian Ye, "Deep Learning Hierarchical Representations for Image Steganalysis", IEEE Transactions on Information Forensics and Security, 12, pp.68970~68990 (2017)
- S. Liu, "Image steganalysis of low embedding rate based on the attention mechanism and transfer learning", MDPI Electronics, 12(4), pp.12~ (2023)
- M. Boroumand, "Deep residual network for steganalysis of digital images", IEEE Transactions on Information Forensics and Security, 14(5), pp.1181~1193 (2018)
- Yang Xu, "DRHNet: A Deep Residual Network Based on Heterogeneous Kernel for Steganalysis", Security and Communication Networks, 2020(1), pp.9~ (2020)
- Yassine Yousfi, "ImageNet Pre-trained CNNs for JPEG Steganalysis", 2020 IEEE International Workshop on Information Forensics and Security (WIFS), pp.6~ (2020)
- Matthew Botvinick, "Review Deep Reinforcement Learning and Its Neuroscientific Implications", 50 Neuron, 107(4), pp.603~616 (2020)
- Jianhua Yang, "Reinforcement Learning Aided Network Architecture Generation for JPEG Image Steganalysis", IH&MMSec, 20, pp.10~ (2020)
- Sanghoon Kang, "Classification of Multiple Steganographic Algorithms Using Hierarchical CNNs and ResNets", SPRING NATURE Link Proceedings of International Conference on Smart Computing and Cyber Security, pp.365~373 (2021)
- Yuting Hu, "Detection of heterogeneous parallel steganography for low bit-rate VoIP speech streams", ScienceDirect Neurcocomputing, 419, pp.70~79 (2021)
- Y. Xu, "Small-scale linguistic steganalysis for multi-concealed scenarios", IEEE Signal Processing Letters, 29, pp.130~134 (2021)
- Weixuan Tang, "CNN-Based Adversarial Embedding for Image Steganography", IEEE Transactions on Information Forensics and Security, 14(8), pp.2074~2087 (2019)
- Lin Li, "AdvSGAN:Adversarial imageSteganography with adversarial networks", SPRING NATURE Link, 80, pp.25539~25555 (2021)
- Nicolas Papernot, "The Limitations of Deep Learning in Adversarial Settings", 2016 IEEE European Symposium on Security and Privacy (EuroS&P), pp.16~ (2016)
- 김민지, "딥러닝 모델에 대한 적대적 예제 공격 기술 동향 연구", 2023년한국산학기술학회 춘계 학술발표논문집, pp.4~ (2023)
- Ian J. Goodfellow, "Generative Adversarial Nets", NeurIPS Proceedings, pp.9~ (2014)
- Dongbin Na, "Unrestricted Black-box Adversarial Attack Using GAN with Limited Queries", SPRING NATURE Link Computer Vision – ECCV 2022 Workshops, pp.467~482 (2023)
- Vida Yousefi Ramandi, "VidaGAN: Adaptive GAN for image steganography", IET Image Processing, 18(12), pp.3329~3342 (2024)
- Tabares-Soto Reinel, "Deep Learning Applied to Steganalysis of Digital Images: A Systematic Review", IEEE Access, 7, pp.68970~68990 (2019)
- 김현재, "딥러닝을 이용한 범용적 스테그아날리시스", KIISE Transactions on Computing Practices, 23(4), pp.244~249
- Nandhini Subramanian, "Image Steganography: A Review of the Recent Advances", IEEE Access, 9, pp.23409~23423 (2021)