Korea Information Processing Society Review (정보처리학회지)

Korea Information Processing Society (한국정보처리학회)
권호 표지

생성 AI 보안 및 프라이버시 이슈

  • Published : 2023.03.31
Download PDF
⟨ Previous Next ⟩

Abstract

Keywords

References

  1. Alec, Radford, et al, "Improving language understanding by generative pre-training.", 2018.
  2. Goodfellow, Ian, et al, "Generative adversarial networks.", Communications of the ACM 63.11, 139-144, 2020. https://doi.org/10.1145/3422622
  3. Ajay Jain, Ho, Jonathan, and Pieter Abbeel, "Denoising diffusion probabilistic models.", Advances in Neural Information Processing Systems 33, pp.6840-6851, 2020.
  4. Robin, Rombach, et al, "High-resolution image synthesis with latent diffusion models.", Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp.10684-10695, 2022.
  5. Alex, Nichol, et al, "Glide: Towards photorealistic image generation and editing with text-guided diffusion models.", arXiv preprint arXiv:2112.10741, 2021.
  6. Chitwan, Saharia, et al, "Photorealistic text-to-image diffusion models with deep language understanding.", arXiv preprint arXiv:2205.11487, 2022.
  7. Diederik P., Max Welling, Kingma, "Auto-encoding variational bayes.", arXiv preprint arXiv:1312.6114, 2013.
  8. Ashish, Vaswani, et al, "Attention is all you need.", Advances in neural information processing systems 30, 2017.
  9. 최선, "의사 시험까지 합격한 chatGPT "과한 기대는 금물"", 2023년 2월 23일자.
  10. 김효영, "ChatGPT 과제 0점처리...철학없는 인공지능은 비교육적", 2023년 2월 20일자.
  11. 공인호, "챗GPT로 작가 등단...저작권 논란 가열", 포춘코리아, 2023년 2월 23일자.
  12. 이상덕, ""나는 살아있다"...공포감 들게 한 소름 돋는 답변의 정체는", 매일경제, 2023년 2월 16일자.
  13. Jun-Yan, Zhu, et al, "Unpaired image-to-image translation using cycle-consistent adversarial networks.", Proceedings of the IEEE international conference on computer vision, pp. 2223-2232, 2017.
  14. Choi, Yunjey, et al, "Stargan: Unified generative adversarial networks for multi-domain image-to-image translation.", Proceedings of the IEEE conference on computer vision and pattern recognition, 2018.
  15. Karras, Samuli Laine, Tero, and Timo Aila, "A style-based generator architecture for generative adversarial networks.", Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp. 4401-4410, 2019.
  16. Jiahui, Yu, et al, "Scaling autoregressive models for content-rich text-to-image generation.", arXiv preprint arXiv:2206.10789, 2022.
  17. Aditya, Ramesh, et al, "Zero-shot text-to-image generation.", International Conference on Machine Learning. PMLR, 2021.
  18. Aditya, Ramesh, et al, "Hierarchical textconditional image generation with clip latents.", arXiv preprint arXiv:2204.06125, 2022.
  19. 김송이, "미술전 1등 작품, ○○가 그렸다고?...논란된 작가, 누구길래", 뉴스1, 2022년 9월 5일자.
  20. 임병선, "김정기 작가 세상 뜨자마자 AI 학습한 그림 올리고 "오마주"", 서울신문, 2022년 10월 10일자.
  21. Bhiksha Raj, Gao, Rita Singh, and Yang,. "Voice impersonation using generative adversarial networks.", 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2018.
  22. Donahue, Chris, Julian McAuley, and Miller Puckette, "Adversarial audio synthesis.", arXiv preprint arXiv:1802.04208, 2018.
  23. Borsos, Zalan, et al, "Audiolm: a language modeling approach to audio generation.", arXiv preprint arXiv:2209.03143, 2022.
  24. Huang, Qingqing, et al, "Noise2Music: Text-conditioned Music Generation with Diffusion Models.", arXiv preprint arXiv:2302.03917, 2023.
  25. 황희진, "아마존 AI '알렉사'의 살인미수? 10살 소녀에 "전기 콘센트에 동전 갖다 대"", 매일신문, 2021년 12월 29일자.
  26. 정경훈, 김창현, "[단독]"목소리 소름주의"...400억 가로챈 '딥보이스 범죄', 檢도 나섰다", 머니투데이, 2023년 2월 11일자.
  27. Ahmed, Salem, et al, "Baaan: Backdoor attacks against autoencoder and gan-based machine learning models.", arXiv preprint arXiv:2010.03007, 2020.
  28. Ambrish, Killian Levacher, Mathieu Sinn, and Rawat, "The Devil Is in the GAN: Backdoor Attacks and Defenses in Deep Generative Models.", Computer Security- ESORICS 2022: 27th European Symposium on Research in Computer Security, Copenhagen, Denmark, September 26-30, 2022, Proceedings, Part III. Cham: Springer Nature Switzerland, pp. 776-783, 2022.
  29. Dominik Hintersdorf, Kristian Kersting, Lukas, and Struppek, "Rickrolling the Artist: Injecting Invisible Backdoors into Text-Guided Image Generation Models.", arXiv preprint arXiv:2211.02408, 2022.
  30. Chou, Pin-Yu Chen, Sheng-Yen, and Tsung-Yi Ho, "How to Backdoor Diffusion Models?.", arXiv preprint arXiv:2212.05400, 2022.
  31. Myle, Ott, et al, "fairseq: A fast, extensible toolkit for sequence modeling.", arXiv preprint arXiv:1904.01038, 2019.
  32. Sun, Xiaofei, et al, "Defending against backdoor attacks in natural language generation.", arXiv preprint arXiv:2106.01810, 2021.
  33. Hailong, Hu, and Jun Pang, "Model extraction and defenses on generative adversarial networks.", arXiv preprint arXiv: 2101.02069, 2021.
  34. Miyato, Takeru, et al, "Spectral normalization for generative adversarial networks.", arXiv preprint arXiv:1802.05957, 2018.
  35. Karras, Tero, et al, "Progressive growing of gans for improved quality, stability, and variation.", arXiv preprint arXiv:1710.10196, 2017.
  36. Hayes, Jamie, et al, "Logan: Membership inference attacks against generative models.", arXiv preprint arXiv:1705.07663, 2017.
  37. Luke Metz, Radford, Alec, and Soumith Chintala, "Unsupervised representation learning with deep convolutional generative adversarial networks.", arXiv preprint arXiv:1511.06434, 2015.
  38. Berthelot, David, Luke Metz, and Thomas Schumm, "Began: Boundary equilibrium generative adversarial networks.", arXiv preprint arXiv:1703.10717, 2017.
  39. Benjamin, Daniel Bernau, Hilprecht, and Martin Harterich, "Monte Carlo and Reconstruction Membership Inference Attacks against Generative Models.", Proc. Priv. Enhancing Technol, 232-249, 2019
  40. Chen, Dingfan, et al, "Gan-leaks: A taxonomy of membership inference attacks against generative models.", Proceedings of the 2020 ACM SIGSAC conference on computer and communications security, pp. 343-362, 2020.
  41. Wu, Yixin, et al, "Membership Inference Attacks Against Text-to-image Generation Models.", arXiv preprint arXiv:2210.00968, 2022.
  42. Duan, Jinhao, et al, "Are Diffusion Models Vulnerable to Membership Inference Attacks?.", arXiv preprint arXiv:2302.01316, 2023.
  43. Carlini, Nicholas, et al, "Extracting training data from diffusion models.", arXiv preprint arXiv:2301.13188, 2023.