EUVIMEDEUVIMED entdecken
Crossref

A Mutual Head Knowledge Distillation Framework for Lightweight RGB-T Crowd Counting

Baoyang Mu, Feng Shao, Hangwei Chen, Xuejin Wang, Qiuping Jiang

IEEE Transactions on Circuits and Systems for Video Technology · 2025-12 · 35 (12) · 12477-12491

DOI: 10.1109/tcsvt.2025.3588815 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Circuits and Systems for Video Technology, 2025-12. Verfasst von Baoyang Mu und 4 weiteren Autor:innen. Die Arbeit wurde bisher 4-mal zitiert (Crossref) und nennt 51 Literaturangaben. Gefördert von: Natural Science Foundation of China, “Science and Technology Innovation Yongjiang 2035” Key Technology Breakthrough Plan Project in Ningbo, Natural Science Foundation of Fujian Province.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

As an important technology in the fields of intelligent transportation and public safety, crowd counting that can obtain pedestrian flow information has attracted extensive attention from academic and industrial communities. However, existing RGB-T crowd counting methods cannot effectively balance the counting accuracy and computational complexity in practical applications. For this, we propose a Mutual Head Knowledge Distillation Framework (MHKDF) to obtain a lightweight RGB-T crowd counting network for efficient and accurate pedestrian number estimation. Specifically, to avoid the influence of parameter and structure differences between teacher and student networks on the distillation effect, we propose a Cooperative Mutual Knowledge Distillation (CMKD) strategy to comprehensively and dynamically transfer the crowd analysis ability of the complex teacher model (MHKDF-T) to the lightweight student model (MHKDF-S). In addition, the upper bound of the performance of the student network depends on the teacher model with high accuracy. Therefore, to take advantage of the complementary advantages of frequency domain and spatial domain feature fusion, we propose a Multi-Modal Spatial-Frequency Hybrid Fusion Module (MSFHFM) to futher improve counting accuracy of MHKDF-T. Comprehensive experiments on two RGB-T crowd counting datasets demonstrate that our MHKDF-S achieves competitive performance with only 5.68 FLOPs and 4.89M parameters. Our code will be released at https://github.com/BaoYangCC/MHKDF

Förderung

  • Natural Science Foundation of China
  • “Science and Technology Innovation Yongjiang 2035” Key Technology Breakthrough Plan Project in Ningbo
  • Natural Science Foundation of Fujian Province

Zitieren

APA

Mu, B., Shao, F., Chen, H., Wang, X., & Jiang, Q. (2025). A Mutual Head Knowledge Distillation Framework for Lightweight RGB-T Crowd Counting. IEEE Transactions on Circuits and Systems for Video Technology, 35(12), 12477-12491. https://doi.org/10.1109/tcsvt.2025.3588815

Vancouver

Mu B, Shao F, Chen H, Wang X, Jiang Q. A Mutual Head Knowledge Distillation Framework for Lightweight RGB-T Crowd Counting. IEEE Transactions on Circuits and Systems for Video Technology. 2025;35(12):12477-12491. doi:10.1109/tcsvt.2025.3588815

BibTeX
@article{Mu2025,
  author = {Mu, Baoyang and Shao, Feng and Chen, Hangwei and Wang, Xuejin and Jiang, Qiuping},
  title = {A Mutual Head Knowledge Distillation Framework for Lightweight RGB-T Crowd Counting},
  journal = {IEEE Transactions on Circuits and Systems for Video Technology},
  year = {2025},
  volume = {35},
  number = {12},
  pages = {12477--12491},
  doi = {10.1109/tcsvt.2025.3588815}
}
Kostenlos starten →