반도체디스플레이기술학회지 (Journal of the Semiconductor & Display Technology)

한국반도체디스플레이기술학회 (The Korean Society Of Semiconductor & Display Technology)
권호 표지

DCT 학습을 융합한 RRU-Net 기반 이미지 스플라이싱 위조 영역 탐지 모델

A DCT Learning Combined RRU-Net for the Image Splicing Forgery Detection

  • 서영민 (상명대학교 정보보안공학과) ;
  • 한정우 (상명대학교 정보보안공학과) ;
  • 권희정 (상명대학교 정보보안공학과) ;
  • 이수빈 (상명대학교 정보보안공학과) ;
  • 국중진 (상명대학교 정보보안공학과)
  • Young-min Seo (Dept. of Information Security Engineering, Sangmyung University) ;
  • Jung-woo Han (Dept. of Information Security Engineering, Sangmyung University) ;
  • Hee-jung Kwon (Dept. of Information Security Engineering, Sangmyung University) ;
  • Su-bin Lee (Dept. of Information Security Engineering, Sangmyung University) ;
  • Joongjin Kook (Dept. of Information Security Engineering, Sangmyung University)
  • 투고 : 2023.01.26
  • 심사 : 2023.03.20
  • 발행 : 2023.03.31
PDF 다운로드
⟨ 이전 논문 다음 논문 ⟩

초록

This paper proposes a lightweight deep learning network for detecting an image splicing forgery. The research on image forgery detection using CNN, a deep learning network, and research on detecting and localizing forgery in pixel units are in progress. Among them, CAT-Net, which learns the discrete cosine transform coefficients of images together with images, was released in 2022. The DCT coefficients presented by CAT-Net are combined with the JPEG artifact learning module and the backbone model as pre-learning, and the weights are fixed. The dataset used for pre-training is not included in the public dataset, and the backbone model has a relatively large number of network parameters, which causes overfitting in a small dataset, hindering generalization performance. In this paper, this learning module is designed to learn the characterization depending on the DCT domain in real-time during network training without pre-training. The DCT RRU-Net proposed in this paper is a network that combines RRU-Net which detects forgery by learning only images and JPEG artifact learning module. It is confirmed that the network parameters are less than those of CAT-Net, the detection performance of forgery is better than that of RRU-Net, and the generalization performance for various datasets improves through the network architecture and training method of DCT RRU-Net.

키워드

참고문헌

  1. D. L. M. Sacchi, F. Agnoli, and E. F. Loftus, "Changing history: doctored photographs affect memory for past public events," vol. 21, no. 8, pp. 1005-1022, Dec. 2007, doi: 10.1002/acp.1394.
  2. M. Mishra and F. L. D. M. C. Adhikary, "Digital Image Tamper Detection Techniques - A Comprehensive Study," Jun. 2013, doi: 10.48550/arxiv.1306.6737.
  3. C. N. Bharti and P. Tandel, "A survey of image forgery detection techniques," 2016, pp. 877-881, doi: 10.1109/WiSPNET.2016.7566257
  4. W. N. Nathalie Diane, S. Xingming, and F. K. Moise, "A Survey of Partition-Based Techniques for CopyMove Forgery Detection," vol. 2014, pp. 975456-13, Jul. 2014, doi: 10.1155/2014/975456.
  5. M. D. Ansari, S. P. Ghrera, and V. Tyagi, "Pixel-Based Image Forgery Detection: A Review," vol. 55, no. 1, pp. 40-46, Jan. 2014, doi: 10.1080/09747338.2014.921415.
  6. G. K. Birajdar and V. H. Mankar, "Digital image forgery detection using passive techniques: A survey," vol. 10, no. 3, pp. 226-245, Oct. 2013, doi: 10.1016/j.diin.2013.04.007.
  7. T. Qazi et al., "Survey on blind image forgery detection," vol. 7, no. 7, pp. 660-670, Oct. 2013, doi: 10.1049/ietipr.2012.0388.
  8. Y. Shi, C. Chen, and W. Chen, "A natural image model approach to splicing detection," 2007, pp. 51-62, doi: 10.1145/1288869.1288878
  9. Z. He, W. Lu, W. Sun, and J. Huang, "Digital image splicing detection based on Markov features in DCT and DWT domain," vol. 45, no. 12, pp. 4292-4299, Dec. 2012, doi: 10.1016/j.patcog.2012.05.014.
  10. S. Velliangiri and J. Premalatha, "A Novel Forgery Detection in Image Frames of the Videos Using Enhanced Convolutional Neural Network in Face Images," vol. 125, no. 2, pp. 625-645, Nov. 2020, doi: 10.32604/cmes.2020.010869.
  11. H. Mo, B. Chen, and W. Luo, "Fake Faces Identification via Convolutional Neural Network," 2018, pp. 43-47, doi: 10.1145/3206004.3206009
  12. Y. Wu, W. AbdAlmageed, and P. Natarajan, "ManTraNet: Manipulation Tracing Network for Detection and Localization of Image Forgeries With Anomalous Features," 2019, pp. 9535-9544, doi: 10.1109/CVPR.2019.00977
  13. X. Bi, Y. Wei, B. Xiao, and W. Li, "RRU-Net: The Ringed Residual U-Net for Image Splicing Forgery Detection," 2019, pp. 30-39, doi: 10.1109/CVPRW.2019.00010
  14. M.-J. Kwon, S.-H. Nam, I.-J. Yu, H.-K. Lee, and C. Kim, "Learning JPEG Compression Artifacts for Image Manipulation Detection and Localization," vol. 130, no. 8, pp. 1875-1895, 2022, doi: 10.1007/s11263-022-01617-5.
  15. J. Wang et al., "Deep High-Resolution Representation Learning for Visual Recognition," vol. 43, no. 10, pp. 3349-3364, Oct. 2021, doi: 10.1109/TPAMI.2020.2983686.org.libproxy.smu.ac.kr/document/9052469
  16. Jia Deng, Wei Dong, R. Socher, Li-Jia Li, Kai Li, and Li Fei-Fei, "ImageNet: A large-scale hierarchical imag e database," 2009, pp. 248-255, doi: 10.1109/CVPR.2009.5206848
  17. Jing Dong, Wei Wang, and Tieniu Tan, "CASIA Image Tampering Detection Evaluation Database," 2013, pp. 422-426, doi: 10.1109/ChinaSIP.2013.6625374
  18. D. P. Kingma and J. Ba, "Adam: A method for stochastic optimization," Dec. 2014, doi: 10.48550/ARXIV.1412.6980.