Empowering Object Detection: Unleashing the Potential of Decoupled and Interactive Distillation
IEEE Transactions on Intelligent Transportation Systems · 2025-01 · 26 (1) · 544-557
DOI: 10.1109/tits.2024.3489888 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Intelligent Transportation Systems, 2025-01. Verfasst von Fulan Qian und 6 weiteren Autor:innen. Die Arbeit wurde bisher 4-mal zitiert (Crossref) und nennt 56 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Scientific Research Planning Project of Anhui Province.
Abstract (Originalsprache)
Deploying state-of-the-art object detectors on resource-limited devices presents significant challenges. Knowledge distillation is an efficient and streamlined lightweight technique to improve the accuracy of compact detectors. However, its effectiveness is limited by the redundancy of different types of semantics on the feature map and the closure of same level’s feature distillation. To alleviate this problem, we propose Decoupled and Interactive Distillation, an effective and versatile method to improve knowledge distillation in some complex object detection tasks. The method has two key components. A knowledge decoupled module captures category awareness and localization awareness features. A multi-level feature interaction distillation can aggregate feature distillations from shallow to deep levels, facilitating the collaboration between feature transfers at different levels. The relevant experiments in traffic-related, 3D, rotated object detection have verified the effectiveness of the proposed method, particularly in challenging scenes.
Förderung
- National Natural Science Foundation of China
- Scientific Research Planning Project of Anhui Province
Zitieren
APA
Qian, F., Hong, J., Yan, H., Chen, H., Zhang, C., Su, H., & Zhao, S. (2025). Empowering Object Detection: Unleashing the Potential of Decoupled and Interactive Distillation. IEEE Transactions on Intelligent Transportation Systems, 26(1), 544-557. https://doi.org/10.1109/tits.2024.3489888
Vancouver
Qian F, Hong J, Yan H, Chen H, Zhang C, Su H, et al. Empowering Object Detection: Unleashing the Potential of Decoupled and Interactive Distillation. IEEE Transactions on Intelligent Transportation Systems. 2025;26(1):544-557. doi:10.1109/tits.2024.3489888
BibTeX
@article{Qian2025,
author = {Qian, Fulan and Hong, Jiacheng and Yan, Huanqian and Chen, Hai and Zhang, Chonghao and Su, Hang and Zhao, Shu},
title = {Empowering Object Detection: Unleashing the Potential of Decoupled and Interactive Distillation},
journal = {IEEE Transactions on Intelligent Transportation Systems},
year = {2025},
volume = {26},
number = {1},
pages = {544--557},
doi = {10.1109/tits.2024.3489888}
}
EUVIMED