Conference paper · 2023
Experimental Face Recognition Using Applied Deep Learning Approaches to Find Missing Persons
Proceedings of International Conference on Frontiers in Computing and Systems: COMSYS 2021, Lecture Notes in Networks and Systems, vol. 404, pp. 3–11, Springer Nature Singapore · Published
Summary
Imoh et al. (2023) built an experimental face-recognition system to help find missing persons in Nigeria, where, according to the International Committee of the Red Cross (2020), a little over 23,000 people were declared missing and existing search methods (word of mouth, media announcements and social media) are described as slow and ineffective. The system combines a convolutional neural network with facial calibration and modelling to extract face encodings, which are compared against a query image to identify a match. The abstract reports no accuracy figures, dataset size or field deployment results.
Practical implication. For Nigerian law enforcement agencies, human rights organisations and families of missing persons, an automated, deep-learning-based face matching tool could speed up search and identification where time is critical.
- The Problem
- More than 23,000 people were reported missing in Nigeria in 2020, and existing search methods — word of mouth, media announcements and social media — are slow and ineffective.
- The Methodology
- Experimental system design: face recognition with a convolutional neural network, facial calibration and modelling for feature extraction, and comparison of face encodings against a query image.
- The Core Finding
- The authors built an experimental face-recognition system that combines a convolutional neural network with facial calibration and modelling to extract face encodings and match them against a query image, as a faster tool for identifying missing persons.
- The Citation
- Imoh, N., Vajjhala, N. R., & Rakshit, S. (2023). Experimental Face Recognition Using Applied Deep Learning Approaches to Find Missing Persons. In S. Basu, D. K. Kole, A. K. Maji, D. Plewczynski, & D. Bhattacharjee (Eds.), Proceedings of International Conference on Frontiers in Computing and Systems: COMSYS 2021 (pp. 3–11). Springer Nature Singapore. https://doi.org/10.1007/978-981-19-0105-8_1
The summary, key findings, methods, implications and relevance notes on this page are this website’s own description of the paper, written from the published abstract and text. The official abstract and citation details are given further down.
Key findings
- Finding 1. According to the ICRC (2020), more than 40,000 people were declared missing in Africa, a little over 23,000 of them in Nigeria; most Nigerian cases are attributed to the insurgency and security crisis of the preceding decade.
- Finding 2. Existing solutions in Nigeria — word of mouth, media and print announcements and, more recently, social media — are described as inefficacious and slow, particularly when time is a determining factor.
- Finding 3. The experimental system combines facial recognition with deep learning using a convolutional neural network, with high-standard facial calibration and modelling used for feature extraction.
- Finding 4. Extracted features form face encodings that are compared against a given image to identify a person.
Source: Imoh et al. (2023), Proceedings of International Conference on Frontiers in Computing and Systems: COMSYS 2021, Lecture Notes in Networks and Systems, vol. 404, pp. 3–11, Springer Nature Singapore. DOI: 10.1007/978-981-19-0105-8_1
The study in detail
What question does this paper answer?
Can a facial recognition system with deep learning functionality help law enforcement agencies, human rights organisations and families find and identify missing persons in Nigeria faster than word of mouth, media announcements and social media?
What did the study find, in detail?
The authors build an experimental system that combines facial recognition with a convolutional neural network. High-standard facial calibration and modelling are used for feature extraction; the extracted features form face encodings that are then compared with a given image to identify a match. The paper motivates the system with ICRC figures of more than 40,000 people reported missing in Africa in 2020, over 23,000 of them in Nigeria, most attributed to insurgency and insecurity.
Why does it matter?
Current approaches to finding missing persons in Nigeria are described as slow and ineffective where time is critical. An automated, deep-learning-based face matching tool could speed up search and identification for law enforcement, human rights organisations and families.
Methods and data
| Problem | Over 40,000 people declared missing in Africa in 2020 (ICRC), more than 23,000 in Nigeria; current search methods are slow and inefficacious |
|---|---|
| Design | Experimental face recognition system |
| Methods | Deep learning with a convolutional neural network; high-standard facial calibration and modelling for feature extraction; face encodings compared with a given image |
| Intended users | Nigerian law enforcement agencies, human rights organisations, friends and families of missing persons |
| Implication | Deep learning face matching could speed up the search-and-find process where time is decisive |
| Citation | Imoh et al. (2023) · DOI 10.1007/978-981-19-0105-8_1 |
Definitions and key terms
- Face encodings
- Features extracted through facial calibration and modelling, which the experimental system compares with a given image to identify a match. — as used in Imoh et al. (2023)
Abstract
The spike in challenges to security as well as information and resource management across the globe has equally borne the rising demand for a better system and technology to curb it. A news release from the International Committee of the Red Cross (ICRC) in 2020 revealed over 40,000 people were declared missing in Africa. A staggering percentage of that number, a little over 23,000, is documented in Nigeria alone. Despite the numerous factors surrounding missing persons globally, at more than 50% of the original figure, it is unsurprising that most of the cases in Nigeria are attributed to the insurgency and security mishap that has plagued the country for almost a decade. Some of the cases remain unsolved for years, causing the victims to remain untraceable, thereby taking up a different identity and existence, especially if they went missing. Current solutions to find missing persons in Nigeria revolve around word of mouth, media and print announcements, and more recently, social media. These solutions are inefficacious, slow, and do not adequately help find and identify missing persons, especially in situations where time is a determining factor. The use of a facial recognition system with deep learning functionality can help Nigerian law enforcement agencies, and other human rights organizations and friends and families of the missing person speed up the search and find process. Our experimental system combines facial recognition with deep learning using a convoluted neural network. In this study, the authors have used high-standard facial calibration and modeling for feature extraction. These extracted features form the face encodings that are after that compared to a given image.
Abstract as published in Proceedings of International Conference on Frontiers in Computing and Systems: COMSYS 2021.
Keywords: Deep learning; Face recognition; Artificial intelligence; Neural network; Machine learning; Convolutional neural networks
Limitations
- The abstract describes an experimental system and its design; it does not report accuracy figures, dataset size or field deployment results.
Practical implications
- According to the authors, a facial recognition system with deep learning functionality can help Nigerian law enforcement agencies, human rights organisations and the friends and families of missing persons speed up the search and find process.
- Families and agencies in Nigeria that rely on word of mouth, media and print announcements or social media are using search methods the paper describes as inefficacious and slow, particularly when time is a determining factor.
- Researchers assessing the experimental face-recognition system will not find accuracy figures, dataset size or field deployment results in the abstract, which describes the system and its design.
Drawn from Imoh et al. (2023), Proceedings of International Conference on Frontiers in Computing and Systems: COMSYS 2021, Lecture Notes in Networks and Systems, vol. 404, pp. 3–11, Springer Nature Singapore.
When this research may be relevant
This paper may be relevant to researchers studying face recognition with convolutional neural networks, humanitarian and law-enforcement applications of computer vision, missing-person identification systems, and technology responses to insecurity and displacement in Nigeria and sub-Saharan Africa.
Research topics addressed: face recognition; deep learning; convolutional neural networks; missing persons; computer vision; facial feature extraction; face encodings; law enforcement technology; Nigeria; humanitarian technology
How to cite
- APA 7
- Imoh, N., Vajjhala, N. R., & Rakshit, S. (2023). Experimental Face Recognition Using Applied Deep Learning Approaches to Find Missing Persons. In S. Basu, D. K. Kole, A. K. Maji, D. Plewczynski, & D. Bhattacharjee (Eds.), Proceedings of International Conference on Frontiers in Computing and Systems: COMSYS 2021 (pp. 3–11). Springer Nature Singapore. https://doi.org/10.1007/978-981-19-0105-8_1
- MLA 9
- Imoh, Nsikak, et al. “Experimental Face Recognition Using Applied Deep Learning Approaches to Find Missing Persons.” Proceedings of International Conference on Frontiers in Computing and Systems: COMSYS 2021, edited by Subhadip Basu et al., Springer Nature Singapore, 2023, pp. 3–11. https://doi.org/10.1007/978-981-19-0105-8_1.
- IEEE
- N. Imoh, N. R. Vajjhala, and S. Rakshit, “Experimental Face Recognition Using Applied Deep Learning Approaches to Find Missing Persons,” in Proceedings of International Conference on Frontiers in Computing and Systems: COMSYS 2021, S. Basu, D. K. Kole, A. K. Maji, D. Plewczynski, and D. Bhattacharjee, Eds. Springer Nature Singapore, 2023, pp. 3–11, doi: 10.1007/978-981-19-0105-8_1.
BibTeX
@inproceedings{imoh2023deep,
title = {Experimental Face Recognition Using Applied Deep Learning Approaches to Find Missing Persons},
author = {Imoh, Nsikak and Vajjhala, Narasimha Rao and Rakshit, Sandip},
booktitle = {Proceedings of International Conference on Frontiers in Computing and Systems: COMSYS 2021},
series = {Lecture Notes in Networks and Systems},
editor = {Basu, Subhadip and Kole, Dipak Kumar and Maji, Arnab Kumar and Plewczynski, Dariusz and Bhattacharjee, Debotosh},
pages = {3--11},
year = {2023},
publisher = {Springer Nature Singapore},
doi = {10.1007/978-981-19-0105-8_1},
url = {https://doi.org/10.1007/978-981-19-0105-8_1}
}RIS
TY - CPAPER TI - Experimental Face Recognition Using Applied Deep Learning Approaches to Find Missing Persons AU - Imoh, Nsikak AU - Vajjhala, Narasimha Rao AU - Rakshit, Sandip ED - Basu, Subhadip ED - Kole, Dipak Kumar ED - Maji, Arnab Kumar ED - Plewczynski, Dariusz ED - Bhattacharjee, Debotosh T2 - Proceedings of International Conference on Frontiers in Computing and Systems: COMSYS 2021 T3 - Lecture Notes in Networks and Systems PY - 2023 DA - 2022/06/28 VL - 404 SP - 3 EP - 11 PB - Springer Nature Singapore SN - 2367-3370 DO - 10.1007/978-981-19-0105-8_1 UR - https://www.narasimharao.net/research/deep-learning-face-recognition-missing-persons-nigeria/ LA - en ER -
Links and files
- DOI: https://doi.org/10.1007/978-981-19-0105-8_1
- Publisher’s page: https://link.springer.com/chapter/10.1007/978-981-19-0105-8_1
- This page as text: https://www.narasimharao.net/research/deep-learning-face-recognition-missing-persons-nigeria/paper.md
- Permanent address of this page: https://www.narasimharao.net/research/deep-learning-face-recognition-missing-persons-nigeria/
Related research
Related publications
Across 92 primary studies (2016–2025), Diffusion Transformers and Gaussian Splatting account for about 36% of studies dated 2024 or later while GAN use has fallen below 3%; training data are demographically skewed and under-documented, and the dominant metrics (PSNR, SSIM, SyncNet) do not measure what deployment requires.
Fine-tuned DINOv2 and EfficientNet-B0 performed best on nine-class DeepWeeds data, but differences among several high-performing models were not consistently statistically significant across random seeds.
A lightweight EfficientNetV2B0 model trained under CLAIM and TRIPOD-AI reporting standards achieved an AUC of 0.967 (95% CI 0.953–0.979) and pneumonia recall of 0.98 on paediatric chest X-rays, missing only about 2% of true cases.
On ten years of daily data for Sofia, Berlin and Tokyo, linear regression matched random forest for temperature at a fraction of the run time and clearly outperformed it for humidity (R² 0.940–0.964), while decision trees performed worst.
A prefetching approach that builds an online navigation graph from server logs and clusters navigation behaviour with WEKA outperformed conventional prefetching techniques on domain client-group logs, particularly once server idle time allowed prefetching across all user predictions.