Deep Learning Based Image Anonymization for Privacy Protection in Social Media

Shakir, Inbithaq; Jabbar, Huda

doi:Under indexing

Proceedings of International Conference on Applied Innovation in IT · 2026/06/12 · Vol. 14 · Issue 3 · pp. 187–196

Deep Learning Based Image Anonymization for Privacy Protection in Social Media

Inbithaq Ahmed Shakir, Huda Abdulaali Abdulbaqi and Asmaa Sadiq Abdul Jabbar

📄 Download PDF DOI: Under indexing

Abstract

Every day, millions of photos get posted to social-media feeds and stored in the cloud, some may even show you in a light that you’re not aware of. Whether it is revealing people’s faces, license plate or merely a few documents in the background that are causing the photo to be online can make thieves or an unauthorized spy get access of some unwanted information about you, and they then use this information for any number nefarious purposes. There are many exiting methods to hide such kind of thing (such as blurring, cropping etc.) Thesauruses are manual works, and not scalable and could let our proposed technique is tackling this giant challenge using smart machine learning that deep. Variant systems can automatically highlight sensitive areas of a digital picture. For instance, it is possible through CNNs to locate things such as text, any face and license plate in an image. That way instead of just blurring out or blocking those areas, we use a GAN to replace them with something that looks legitimate, keeping the image human and aesthetically pleasing as well, We evaluated our approach on real photographs extracted from genuine social media posts, as well as traditional benchmark datasets, and the results are quite promising. It not only requires no processing after finished acquisition but also has an extremely high detection accuracy rate of 96% and better for young women with somewhat blundered images of good quality. That is what makes it a practical, scalable tool for protecting individuals’ privacy on today’s grand social media plateaus.

Keywords

Visual Privacy Deep Learning Image Anonymization Social Media CNN GAN Anonymization.

References

L. Laishram, S. Sharma, A. Kumar, and R. Singh, “Toward a privacy-preserving face recognition system: A survey of leakages and solutions,” ACM Comput. Surv., vol. 57, no. 6, pp. 1-47, Feb. 2025, [Online]. Available: https://doi.org/10.1145/3673224.
M. Akram, A. Zainab, and H. Khan, “Innovative deep learning image technologies: Applications of deep learning in image processing,” in Modern Intelligent Techniques for Image Processing, pp. 36-55, 2023, [Online]. Available: https://doi.org/10.4018/979-8-3693-9045-0.ch007.
J. Tekli, B. Al Bouna, G. Tekli, and R. Couturier, “A framework for evaluating image obfuscation under deep learning-assisted privacy attacks,” Multimedia Tools Appl., vol. 82, no. 27, pp. 1-33, Apr. 2023, [Online]. Available: https://doi.org/10.1007/s11042-023-14664-y.
S. M. Shvai, V. T. M. H. et al., “Adaptive image anonymization in the context of image classification with privacy preservation,” in Proc. ICCV, 2023.
A. Ciftci, I. Demir, and L. Yin, “My Face My Choice: Privacy Enhancing Deepfakes for Social Media,” in Proc. WACV, 2023, doi: 10.1109/WACV56688.2023.00142.
S. Shirai and J. Whitehill, “Privacy-preserving annotation of face images through attribute-preserving face synthesis,” in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. Workshops, 2019.
S. Ren, K. He, R. Girshick, and J. Sun, “Faster R-CNN: Towards real-time object detection with region proposal networks,” in Adv. Neural Inf. Process. Syst. (NeurIPS), 2015, [Online]. Available: https://arxiv.org/abs/1506.01497.
S. Sun, M. Zhao, and Y.-G. Jiang, “Natural and effective obfuscation by head inpainting,” in Proc. Eur. Conf. Comput. Vis. (ECCV), 2018.
C. Liu, X. Wu, Y. Li, and H. Zhang, “Privacy intelligence: A survey on image privacy in online social networks,” ACM Comput. Surv., vol. 55, no. 8, art. 161, pp. 1-35, Dec. 2022, [Online]. Available: https://doi.org/10.1145/3547299.
Y. LeCun, Y. Bengio, and G. Hinton, “Deep learning,” Nature, vol. 521, no. 7553, pp. 436-444, 2015, [Online]. Available: https://doi.org/10.1038/nature14539.
T.-N. Pham, V.-H. Nguyen, K.-R. Kwon, J.-H. Kim, and J.-H. Huh, “Improved YOLOv5 based deep learning system for jellyfish detection,” IEEE Access, 2024, [Online]. Available: https://doi.org/10.1109/ACCESS.2024.3405452.
X. Sun, X. Peng, and J. Xu, “GAN-based visual privacy protection using real-time face detection and anonymization,” IEEE Access, vol. 9, pp. 16245-16258, 2021, [Online]. Available: https://doi.org/10.1109/ACCESS.2021.3053393.
K. Huang, Y. Chen, J. Lin, and C. Shen, “A lightweight privacy-preserving CNN feature extraction framework for mobile sensing,” IEEE Trans. Dependable Secure Comput., vol. 18, no. 3, pp. 1441-1455, 2021, [Online]. Available: https://doi.org/10.1109/TDSC.2019.2913362.
I. Goodfellow et al., “Generative adversarial nets,” in Adv. Neural Inf. Process. Syst. (NeurIPS), 2014.
M. A. Souibgui and Y. Kessentini, “DE-GAN: A conditional generative adversarial network for document enhancement,” IEEE Trans. Pattern Anal. Mach. Intell., 2020, [Online]. Available: https://doi.org/10.1109/TPAMI.2020.3022406.
X. Sun, X. Peng, and J. Xu, “Real-time face anonymization using GANs with identity obfuscation,” IEEE Access, vol. 9, pp. 54735-54745, 2021, [Online]. Available: https://doi.org/10.1109/ACCESS.2021.3070294.
W. Yang, Z. Liu, and D. Zhan, “PrivacyGAN: Protecting facial privacy via adversarial image replacement,” Pattern Recognit. Lett., vol. 155, pp. 37-44, 2022, [Online]. Available: https://doi.org/10.1016/j.patrec.2021.12.005.
Y. Huang, J. Wang, and F. Liu, “Context-aware anonymization using semantic cGANs for privacy-preserving surveillance,” Sensors, vol. 22, no. 8, p. 2951, 2022, [Online]. Available: https://doi.org/10.3390/s22082951.
L. Qi, C. Zhang, and X. Chen, “cGAN-based facial anonymization with perceptual preservation,” Comput. Secur., vol. 125, p. 102957, 2023, [Online]. Available: https://doi.org/10.1016/j.cose.2022.102957.
C. Zhang, J. Liu, T. Wang, and X. He, “A privacy-preserving multi-task learning framework for face detection, landmark localization, pose estimation, and gender recognition,” Front. Neurorobot., vol. 13, art. 112, 2020, [Online]. Available: https://doi.org/10.3389/fnbot.2019.00112.
Y. Zhang and Q. Yang, “A survey on multi-task learning,” IEEE Trans. Knowl. Data Eng., vol. 34, no. 12, pp. 5586-5609, 2022, [Online]. Available: https://doi.org/10.1109/TKDE.2021.3076624.
R. Archana and P. S. E. Jeevara, “Deep learning models for digital image processing,” Artif. Intell. Rev., vol. 57, art. 11, pp. 2-33, 2024, [Online]. Available: https://doi.org/10.1007/s10462-023-10631.
K. Zhang, Z. Zhang, Z. Li, and Y. Qiao, “Joint face detection and alignment using multitask cascaded convolutional networks,” IEEE Signal Process. Lett., vol. 23, no. 10, pp. 1499-1503, 2016, [Online]. Available: https://doi.org/10.1109/LSP.2016.2603342.
S. Tekli, B. Al Bouna, G. Tekli, and R. Couturier, “A framework for evaluating image obfuscation under deep learning-assisted privacy attacks,” Multimedia Tools Appl., vol. 82, no. 27, pp. 1-33, 2023, doi: 10.1109/PST47121.2019.8949040.
Kuznetsova, H. Rom, N. Alldrin, J. Uijlings, and V. Ferrari, “The Open Images dataset V4: Unified image classification, object detection, and visual relationship detection at scale,” Int. J. Inf. Technol., vol. 128, pp. 1956-1981, 2020.
T. Saidani, A. Cheikh, and M. Boussaïd, “Deep learning approach: YOLOv5-based custom object detection,” Eng. Technol. Appl. Sci. Res., vol. 13, no. 6, pp. 12158-12163, 2023, [Online]. Available: https://doi.org/10.48084/etasr.6397.
S. Babitha Rani et al., “Human detection and counting using YOLOv4,” J. Electr. Syst., vol. 20, no. 9s, 2024, [Online]. Available: https://doi.org/10.52783/jes.4945.
J. Kniežová, P. Hrkút, and E. Kršák, “Data pseudonymization in the generative artificial intelligence environment,” in 37th Conf. Open Innovations Assoc. (FRUCT), 2025, [Online]. Available: https://doi.org/10.23919/FRUCT65909.2025.11008283.