P2CODE
Published 2025 | Version v1
Preprint Open

Impact of Face Detection Algorithms on UAV-Based Real-Time Face Recognition Systems

Authors/Creators

Description

Unmanned Aerial Vehicles (UAVs) equipped with real-time face recognition systems play a crucial role in security and surveillance applications, particularly for intruder detection in high-security environments. However, the efficiency of these systems heavily depends on the performance of the face detection algorithm, which serves as the first stage in the face recognition pipeline.

This paper evaluates the impact of three state-of-the-art face detection algorithms—UWS-YOLO, YOLOv7, and RetinaFace—on the real-time performance of a UAV-based face recognition system. Experimental results show that UWS-YOLO achieves the best balance between speed and accuracy, with an inference time of 12.5 ms, an Intruder Detector AI processing time of 26.8 ms, and an overall face recognition pipeline execution time of 48.3 ms, outperforming YOLOv7 (59.3 ms) and RetinaFace (67.9 ms).

These findings highlight the importance of minimising face detection latency to ensure real-time UAV-based surveillance operations.

Files

Impact of Face Detection Algorithms on UAV Face Recognition.pdf

Files (3.7 MB)

Additional details

Funding

European Commission
INCODE - Programming Platform for Intelligent Collaborative Deployments over Heterogeneous Edge-IoT Environments 101093069

References

  • [1] J. Diez-Tomillo, J. Saez-Perez, I. Martinez-Alpiste, C. Lessi, T. Xirofotos, K. Athanasios, D. Klonidis, J. M. Alcaraz-Calero, and Q. Wang, "Next-gen Industry 4.0 with 5G: Enabling secure and high-performance services for critical infrastructure," in 2024 7th International Balkan Conference on Communications and Networking (BalkanCom). IEEE, 2024, pp. 249–253. [2] EU Horizon P2CODE Project, "Programming Platform for Intelligent Collaborative Deployments over Heterogeneous Edge-IoT Environments," Available: https://p2code-project.eu/. [3] J. Diez-Tomillo, I. Martinez-Alpiste, G. Golcarenarenji, Q. Wang, and J. M. Alcaraz-Calero, "Efficient CNN-based low-resolution facial detection from UAVs," Neural Computing and Applications, vol. 36, no. 11, pp. 5847–5860, 2024. [4] C.-Y. Wang, A. Bochkovskiy, and H.-Y. M. Liao, "YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors," in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 7464–7475. [5] J. Deng, J. Guo, Y. Zhou, J. Yu, I. Kotsia, and S. Zafeiriou, "RetinaFace: Single-stage dense face localisation in the wild," 2019. [6] P. Viola and M. Jones, "Rapid object detection using a boosted cascade of simple features," in Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR), vol. 1, 2001, pp. I–511–I–518. [7] H. A. Rowley, S. Baluja, and T. Kanade, "Neural network-based face detection," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 20, no. 1, pp. 23–38, 1998. [8] W. Chen, H. Huang, S. Peng, C. Zhou, and C. Zhang, "YOLO-Face: A real-time face detector," The Visual Computer, vol. 37, pp. 805–813, 2021. [9] Z. Yu, H. Huang, W. Chen, Y. Su, Y. Liu, and X. Wang, "YOLO-FaceV2: A scale and occlusion-aware face detector," Pattern Recognition, vol. 155, p. 110714, 2024. [10] W. Wu, H. Peng, and S. Yu, "YuNet: A tiny millisecond-level face detector," Machine Intelligence Research, vol. 20, no. 5, pp. 656–665, 2023. [11] D. Qi, W. Tan, Q. Yao, and J. Liu, "YOLO5Face: Why reinventing a face detector," in European Conference on Computer Vision. Springer, 2022, pp. 228–244. [12] G. Wang, J. Li, Z. Wu, J. Xu, J. Shen, and W. Yang, "EfficientFace: An efficient deep network with feature enhancement for accurate face detection," Multimedia Systems, vol. 29, no. 5, pp. 2825–2839, 2023. [13] S. Yang, P. Luo, C.-C. Loy, and X. Tang, "WIDER Face: A face detection benchmark," in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2016, pp. 5525–5533. [14] T.-Y. Lin et al., "Microsoft COCO: Common objects in context," in Computer Vision – ECCV 2014. Springer, 2014, pp. 740–755. [15] J. Diez-Tomillo, J. M. Alcaraz-Calero, and Q. Wang, "Dynamic distance-based thresholding for UAV-based face verification algorithms," Sensors, vol. 23, no. 24, p. 9909, 2023. [16] Bluenviron, "MediaMTX: Ready-to-use SRT / WebRTC / RTSP / RTMP / LL-HLS media server and media proxy," 2023. Available: https://github.com/bluenviron/mediamtx [17] Erlang Solutions Limited, "RabbitMQ: A messaging broker software," 2023. Available: https://www.rabbitmq.com/ [18] Weaviate, "Weaviate: An open-source vector embedding search engine," 2023. Available: https://www.weaviate.io/ [19] G. Bradski and A. Kaehler, "OpenCV," Dr. Dobb's Journal of Software Tools, 2000. [20] H. Schulzrinne et al., "Real-Time Streaming Protocol Version 2.0," RFC 7826, Dec. 2016. Available: https://datatracker.ietf.org/doc/rfc7826/ [21] Adobe Systems Incorporated, "Real-Time Messaging Protocol," Tech. Rep., 2012. Available: https://rtmp.veriskope.com/docs/spec/ [22] J. Deng, J. Guo, N. Xue, and S. Zafeiriou, "ArcFace: Additive angular margin loss for deep face recognition," in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2019, pp. 4690–4699. [23] B. Koonce, "ResNet-50," in Convolutional Neural Networks with Swift for TensorFlow: Image Recognition and Dataset Categorisation, 2021, pp. 63–72. [24] A. Bochkovskiy, C.-Y. Wang, and H.-Y. M. Liao, "YOLOv4: Optimal speed and accuracy of object detection," arXiv preprint arXiv:2004.10934, 2020. [25] J. Diez-Tomillo, J. M. Alcaraz-Calero, and Q. Wang, "Face verification algorithms for UAV applications: An empirical comparative analysis," Journal of Communications Software and Systems, vol. 20, no. 1, pp. 1–12, 2024. [26] DJI, "DJI Mini 2 – Small but mighty drone," 2020. Available: https://www.dji.com/mini-2