EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

LittleFaceNet: A Small-Sized Face Recognition Method Based on RetinaFace and AdaFace

Zhengwei Ren, Xinyu Liu, Jing Xu, Yongsheng Zhang, Ming Fang

Journal of Imaging · 13.01.2025 · 11 (1) · 24

DOI: 10.3390/jimaging11010024 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Imaging, online erschienen am 13.01.2025. Verfasst von Zhengwei Ren und 4 weiteren Autor:innen. Beteiligte Einrichtung: School of Artificial Intelligence, Changchun University of Science and Technology, Changchun 130012, China. Die Arbeit wurde bisher 12-mal zitiert (Crossref) und nennt 43 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Science and Technology Research Project of Jilin Provincial Department of Education, Research on the Intelligent Management Mode and Technology of Open Laboratories in Colleges and Universities under the background of Collaborative Innovation o…, Jilin Provincial Education Science Planning Project.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

For surveillance video management in university laboratories, issues such as occlusion and low-resolution face capture often arise. Traditional face recognition algorithms are typically static and rely heavily on clear images, resulting in inaccurate recognition for low-resolution, small-sized faces. To address the challenges of occlusion and low-resolution person identification, this paper proposes a new face recognition framework by reconstructing Retinaface-Resnet and combining it with Quality-Adaptive Margin (adaface). Currently, although there are many target detection algorithms, they all require a large amount of data for training. However, datasets for low-resolution face detection are scarce, leading to poor detection performance of the models. This paper aims to solve Retinaface’s weak face recognition capability in low-resolution scenarios and its potential inaccuracies in face bounding box localization when faces are at extreme angles or partially occluded. To this end, Spatial Depth-wise Separable Convolutions are introduced. Retinaface-Resnet is designed for face detection and localization, while adaface is employed to address low-resolution face recognition by using feature norm approximation to estimate image quality and applying an adaptive margin function. Additionally, a multi-object tracking algorithm is used to solve the problem of moving occlusion. Experimental results demonstrate significant improvements, achieving an accuracy of 96.12% on the WiderFace dataset and a recognition accuracy of 84.36% in practical laboratory applications.

Schlagwörter

Stichwörter der Autor:innen

AdafaceRetinaFacedeep learningface recognitionlow resolutionobject tracking

Autor:innen und Institutionen

  • School of Artificial Intelligence, Changchun University of Science and Technology, Changchun 130012, China.
  • Zhongshan Institute of Changchun University of Science and Technology, Zhongshan 528400, China.

Förderung

  • Science and Technology Research Project of Jilin Provincial Department of Education
  • Research on the Intelligent Management Mode and Technology of Open Laboratories in Colleges and Universities under the background of Collaborative Innovation o…
  • Jilin Provincial Education Science Planning Project
  • Research and Innovation Team Project of Zhongshan Science and Technology Bureau

Zitieren

APA

Ren, Z., Liu, X., Xu, J., Zhang, Y., & Fang, M. (2025). LittleFaceNet: A Small-Sized Face Recognition Method Based on RetinaFace and AdaFace. Journal of Imaging, 11(1), 24. https://doi.org/10.3390/jimaging11010024

Vancouver

Ren Z, Liu X, Xu J, Zhang Y, Fang M. LittleFaceNet: A Small-Sized Face Recognition Method Based on RetinaFace and AdaFace. Journal of Imaging. 2025;11(1):24. doi:10.3390/jimaging11010024

BibTeX
@article{Ren2025,
  author = {Ren, Zhengwei and Liu, Xinyu and Xu, Jing and Zhang, Yongsheng and Fang, Ming},
  title = {LittleFaceNet: A Small-Sized Face Recognition Method Based on RetinaFace and AdaFace},
  journal = {Journal of Imaging},
  year = {2025},
  volume = {11},
  number = {1},
  pages = {24},
  doi = {10.3390/jimaging11010024}
}
Kostenlos starten →