IRO Journals
Journal of Information Technology and Digital World

VeriSphere: A Privacy-Preserving Hybrid Framework for Real-Time Aadhaar Document Fraud Detection

Open Access
Published: 20 August, 2026
Pages: 238-253
  • Shaurya Dawra , Shaurya Dawra
    Department of Information Technology, SRM Institute of Science and Technology, Ramapuram, Chennai, India
    Department of Information Technology, SRM Institute of Science and Technology, Ramapuram, Chennai, India
  • Purvi Mohanty , Purvi Mohanty
    Department of Information Technology, SRM Institute of Science and Technology, Ramapuram, Chennai, India
    Department of Information Technology, SRM Institute of Science and Technology, Ramapuram, Chennai, India
  • Dheeraj Singha Dheeraj Singha
    Department of Information Technology, SRM Institute of Science and Technology, Ramapuram, Chennai, India
    Department of Information Technology, SRM Institute of Science and Technology, Ramapuram, Chennai, India
view PDF
view PDF

How to Cite

Dawra, Shaurya, Purvi Mohanty, and Dheeraj Singha. 2026. “VeriSphere: A Privacy-Preserving Hybrid Framework for Real-Time Aadhaar Document Fraud Detection”. Journal of Information Technology and Digital World 8 (3): 238-53. https://doi.org/10.36548/jitdw.2026.3.008.
ACM ACS APA ABNT Chicago Harvard IEEE MLA Turabian Vancouver AMA

Download Citation

Endnote/Zotero/Mendeley (RIS) BibTeX

Keywords

Aadhaar Document Fraud Detection
Multi-Modal Verification
Decision-Level Fusion
Digital Image Forensics
Verhoeff Checksum Validation
Privacy-Preserving Authentication

Abstract

This research work presents a solution to the increasing problem of Aadhaar document forgery, which poses challenges in ensuring secure identification, in the form of VeriSphere, a complete multi-modal solution for real-time fraud detection. This framework includes a wide variety of methods such as object detection, optical character recognition (OCR), cryptography and biometrics into a unified decision-making architecture. Object detection is carried out using YOLOv8, which identifies the relevant fields in the document. EasyOCR, along with multi head attention mechanism, is used for performing text extraction from these fields. The validity of the Aadhaar number is checked using the Verhoeff algorithm. QR Code decoding and SSIM based matching provide cross modality verification. Further, other methods like Error Level Analysis (ELA) and ResNet-18 are utilized for detecting the artifacts of tampering. All the verification modules' outcomes are integrated using weighted fusion of fraud scores to reach an authenticity decision. The experimental results proved that VeriSphere offers a 97.3% accuracy, 96.8% precision, 98.1% recall and 97.4% F1-score.

References

  1. Mahfoudi, Gaël, Frédéric Morain-Nicolier, Florent Retraint, and Marc Pic. ”CMID: A New Dataset for Copy-Move Forgeries on ID Documents.” IEEE International Conference on Image Processing (ICIP) 2021: 3028-3032.
  2. Zhao, Lin, Changsheng Chen, and Jiwu Huang. ”Deep Learning-Based Forgery Attack on Document Images.” IEEE Transactions on Image Processing 2021, vol 30: 7964-7979.
  3. Fridrich, Jessica, David Soukal, and Jan Lukas. ”Detection of Copy-Move Forgery in Digital Images.” In Proceedings of Digital Forensic Research Workshop 2003, vol. 3, no. 2: 652-63.
  4. Wang, Zhou, Alan C. Bovik, Hamid R. Sheikh, and Eero P. Simoncelli. ”Image Quality Assessment: From Error Visibility to Structural Similarity.” IEEE Transactions on Image Processing 2004, vol. 13, no. 4: 600-612.
  5. He, Kaiming, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. ”Deep Residual Learning for Image Recognition.” In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2016: 770-778.
  6. Redmon, Joseph, and Ali Farhadi. ”Yolov3: An Incremental Improvement.” arXiv preprint 2018, arXiv:1804.02767.
  7. Shi, Baoguang, Xiang Bai, and Cong Yao. ”An End-To-End Trainable Neural Network for Image-Based Sequence Recognition and its Application to Scene Text Recognition.” IEEE Transactions on Pattern Analysis and Machine Intelligence 2016, vol. 39, no. 11:2298-2304.
  8. Xu, Jianjun, Yuxin Wang, Hongtao Xie, and Yongdong Zhang. ”OTE: Exploring Accurate Scene Text Recognition Using One Token.” In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition 2024: 28327-28336.
  9. Vaswani, Ashish, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Łukasz Kaiser, and Illia Polosukhin. ”Attention is All You Need.” Advances in Neural Information Processing Systems 2017, vol. 30: 1-11.
  10. Raganato, Alessandro, and Jörg Tiedemann. ”An Analysis of Encoder Representations in Transformer-Based Machine Translation.” In Proceedings of the 2018 EMNLP Workshop BlackboxNLP: Analyzing and Interpreting Neural Networks for NLP: 287-297.
  11. Varun Kumar Gera, Aadhar Images Dataset, Kaggle, 2023. Available: https://www.kaggle.com/datasets/varunkumargera/aadhar-images