Copyright © 2026 Authors retain the copyright of this article. This article 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.

@article{198468,
        author = {Mr.K. Narsimhulu and Gannamaneni Srihitha and Vadde UmaMaheshwari and Kejiya Eslavath},
        title = {AI BASED MISSING PERSON RECOGNITION SYSTEM WITH MTCNN},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {11},
        pages = {11448-11455},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=198468},
        abstract = {Missing persons identification constitutes one of the major challenges facing society today, and thus there is a need for efficient and scalable solutions. This paper discusses the development of a web-based intelligent system for detecting missing persons through face recognition methods. The developed solution employs artificial intelligence methods such as deep neural network methods and interfaces to help in the identification of missing individuals. To enable the developed system to detect missing individuals' facial features, it utilizes a Multi-task Cascaded Convolutional Neural Network method for facial detection and alignment as well as the InceptionResnetV1 technique to generate the 512-dimensional embeddings of the faces. The faces will then be normalized and kept in a database that consists of faces of both missing and found individuals. Face recognition is done using the match based on the cosine similarity score using a threshold of similarity that is usually between 0.7 and 0.8.
In addition to the facial characteristics of missing individuals, other metadata such as names, age, gender, and places are incorporated into the system to help fine-tune the search process and increase its reliability. Some of the functionalities provided by the platform include registering missing and found individuals, automatic matching, and a search portal from which searches can be done using either names or places or even by uploading images.
The proposed solution is designed to minimize the work involved in the search operation but still locate more people. While the output produced by the solution is impressive under lab tests, there are several environmental variables that determine its success, with image clarity being one of them. Further improvements in the solution can be attained through techniques such as large-scale testing and alerts through instant messages.},
        keywords = {Missing Person Detection, Face Recognition, MTCNN, InceptionResnetV1, Cosine Similarity, Deep Learning, Streamlit.},
        month = {April},
        }