Cascaded Cross-Domain Interaction Network for Video Crowd Counting
IEEE Transactions on Circuits and Systems for Video Technology · 2026-08 · 36 (8) · 12572-12585
DOI: 10.1109/tcsvt.2026.3685264 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Circuits and Systems for Video Technology, 2026-08. Verfasst von Miaogen Ling und 3 weiteren Autor:innen. Die Arbeit nennt 77 Literaturangaben. Gefördert von: Jiangsu Science Foundation, National Science Foundation of China, New Cornerstone Science Foundation through the XPLORER PRIZE.
Abstract (Originalsprache)
Most existing video crowd counting methods focus on spatiotemporal correlations modeling to obtain more accurate and robust crowd number estimations. However, relying solely on the spatiotemporal domain limits the feature representation capability in complex scenes. In comparison to spatiotemporal domain features, frequency-domain features exhibit greater robustness to noise and illumination variations, while reducing feature redundancy. Therefore, in this paper, we propose a cascaded cross-domain feature interaction network that combines spatial and temporal-related frequency-domain features for video crowd counting. First, a unified representation of the frequency-domain feature for each frame is obtained by integrating the high- and low-frequency signals from Haar wavelet transform. Then, a bi-directional channel-wise cross attention is proposed to construct a temporal-related frequency representation of each frame. To obtain more robust spatial and frequency-domain features, a lightweight frequency-space mutual modulation method is proposed to enhance the representation capability of each domain feature. On the Venice dataset, CCINet achieves 5.4%/10.4% improvements in MAE/RMSE over the best existing method. The code is available at: https://github.com/sunjunyu777/CCINet
Förderung
- Jiangsu Science Foundation
- National Science Foundation of China
- New Cornerstone Science Foundation through the XPLORER PRIZE
- Fundamental Research Funds for the Central Universities
- Postgraduate Research & Practice Innovation Program of Jiangsu Province
Zitieren
APA
Ling, M., Sun, J., Fang, W., & Geng, X. (2026). Cascaded Cross-Domain Interaction Network for Video Crowd Counting. IEEE Transactions on Circuits and Systems for Video Technology, 36(8), 12572-12585. https://doi.org/10.1109/tcsvt.2026.3685264
Vancouver
Ling M, Sun J, Fang W, Geng X. Cascaded Cross-Domain Interaction Network for Video Crowd Counting. IEEE Transactions on Circuits and Systems for Video Technology. 2026;36(8):12572-12585. doi:10.1109/tcsvt.2026.3685264
BibTeX
@article{Ling2026,
author = {Ling, Miaogen and Sun, Junyu and Fang, Wei and Geng, Xin},
title = {Cascaded Cross-Domain Interaction Network for Video Crowd Counting},
journal = {IEEE Transactions on Circuits and Systems for Video Technology},
year = {2026},
volume = {36},
number = {8},
pages = {12572--12585},
doi = {10.1109/tcsvt.2026.3685264}
}
EUVIMED