Criminal Face Sketch Recognition and Construction

Abstract

Let’s face it—hand-drawn facial sketches in forensic science just don’t cut it anymore. They take ages to finish, and honestly, they don’t always play nice with the tech tools police actually use to recognize suspects. That’s why this paper introduces a new standalone app. With it, anyone can whip up a composite sketch of a suspect—no need to call in a professional forensic artist. You just drag and drop facial features together, and the app does the rest. Once you’ve built a face, the system automatically checks it against police records. Thanks to deep learning and a cloud setup, this whole process runs way faster and gets better results, giving law enforcement a real boost when it comes to tracking down suspects.

Introduction

The document discusses the evolution and challenges of forensic face sketching used in criminal investigations. Traditional hand-drawn sketches and early digital methods were slow, inaccurate, and often unreliable due to limited feature options and poor realism. To address these issues, modern systems now use AI-powered applications that allow users to create or upload facial features and sketches, improving accuracy and usability.

A review of existing research shows common limitations such as lack of real-world datasets, poor validation, limited scalability, and insufficient integration of advanced AI techniques. Current systems still depend heavily on manual input, struggle with low-quality or varied sketches, and lack real-time database integration.

The proposed system aims to overcome these gaps by developing a standalone application with a drag-and-drop interface, deep learning-based recognition, and cloud integration. It enables realistic sketch creation, automated feature suggestions, and fast matching with police databases. Security is enhanced through features like OTP verification and restricted access.

The methodology combines sketch construction and recognition using machine learning, computer vision, and cloud-based infrastructure. The system is designed to be efficient, accurate, and secure.

Expected outcomes include faster suspect identification, over 90% accuracy, reduced reliance on sketch artists, improved investigation speed, and scalability for future enhancements like CCTV integration and 3D mapping.

Overall, the system significantly improves upon traditional and existing methods by offering a reliable, AI-driven solution for modern forensic investigations.

References

[1] Y. Guo, L. Cao, C. Chen, K. Du and C. Fu, \"Domain Alignment Embedding Network for Sketch Face Recognition,\" in IEEE Access, vol. 9, pp. 872-882, 2021, doi: 10.1109/ACCESS.2020.3047108. [2] Rubeena and E. Kavitha, \"Sketch face Recognition using Deep Learning,\" 2021 5th International Conference on Electronics, Communication and Aerospace Technology (ICECA), Coimbatore, India, 2021, pp. 928-930, doi: 10.1109/ICECA52323.2021.9676049. [3] J. Zheng, W. Song, Y. Wu, R. Xu and F. Liu, \"Feature Encoder Guided Generative Adversarial Network for Face Photo-Sketch Synthesis,\" in IEEE Access, vol. 7, pp. 154971-154985, 2019, doi: 10.1109/ACCESS.2019.2949070. [4] C. Galea and R. A. Farrugia, \"Forensic Face Photo-Sketch Recognition Using a Deep Learning-Based Architecture,\" in IEEE Signal Processing Letters, vol. 24, no. 11, pp. 1586-1590, Nov. 2017, doi: 10.1109/LSP.2017.2749266 [5] L. Wang, V. Sindagi and V. Patel, \"High-Quality Facial Photo-Sketch Synthesis Using Multi-Adversarial Networks,\" 2018 13th IEEE International Conference on Automatic Face & Gesture Recognition (FG 2018), Xi\'an, China, 2018, pp. 83-90, doi: 10.1109/FG.2018.00022. [6] K. Du, H. Zhou, L. Cao, Y. Guo and T. Wang, \"MHGAN: Multi-Hierarchies Generative Adversarial Network for High-Quality Face Sketch Synthesis,\" in IEEE Access, vol. 8, pp. 212995-213011, 2020, doi: 10.1109/ACCESS.2020.3041284. [7] S. M. Iranmanesh, H. Kazemi, S. Soleymani, A. Dabouei and N. M. Nasrabadi, \"Deep Sketch-Photo Face Recognition Assisted by Facial Attributes,\" 2018 IEEE 9th International Conference on Biometrics Theory, Applications and Systems (BTAS), Redondo Beach, CA, USA, 2018, pp. 1-10, doi: 10.1109/BTAS.2018.8698564. [8] Xiaoou Tang and Xiaogang Wang, \"Face sketch synthesis and recognition,\" Proceedings Ninth IEEE International Conference on Computer Vision, Nice, France, 2003, pp. 687-694 vol.1, doi: 10.1109/ICCV.2003.1238414. [9] B. Sheng, P. Li, C. Gao and K. -L. Ma, \"Deep Neural Representation Guided Face Sketch Synthesis,\" in IEEE Transactions on Visualization and Computer Graphics, vol. 25, no. 12, pp. 3216-3230, 1 Dec. 2019, doi: 10.1109/TVCG.2018.2866090. [10] M. Hassan, M. M. Iqbal, H. Qayyum, M. Nawaz, and F. Ali, “Sketch-Based Face Recognition Using Deep Learning,” Webology, vol. 19, no. 3, pp. 3790–3796, 2022. ISSN: 1735-188X. [11] K. Kusumanjali, P. A. Srinivas, M. Thanuja, M. B. Sri, K. V. Priya, and V. Ramarao, “Forensic Sketch Reconnaissance Using Deep Learning,” International Journal of All Research Education and Scientific Methods (IJARESM), vol. 10, no. 5, pp. 3341–3345, May 2022. ISSN: 2455-6211. [12] A. Rasal et al., “Advanced Forensic Face Sketching and Recognition,” International Research Journal of Engineering and Technology (IRJET), vol. 11, no. 4, 2024. ISSN: 2395-0072. [13] A. Rasal et al., “FaceTrace AI: AI-Driven Forensic Sketching and Recognition,” International Journal of Innovative Research in Science, Engineering and Technology (IJIRSET), vol. 12, no. 3, 2025. ISSN: 2347-6710. [14] “Deep Learning-Based Automated Face Sketch Creation and Recognition,” International Journal of Innovative Science and Research Technology (IJISRT), vol. 9, no. 11, Nov. 2024. DOI: 10.5281/zenodo.10459897

Copyright

Copyright © 2026 Hardik Sanap, Aditi Rasal, Shreya Dadhekar, Leena Raut. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.