@Article{cmc.2024.048928,
AUTHOR = {Mingze Li, Diwen Zheng, Shuhua Lu},
TITLE = {Lightweight Res-Connection Multi-Branch Network for Highly Accurate Crowd Counting and Localization},
JOURNAL = {Computers, Materials \& Continua},
VOLUME = {79},
YEAR = {2024},
NUMBER = {2},
PAGES = {2105--2122},
URL = {http://www.techscience.com/cmc/v79n2/56427},
ISSN = {1546-2226},
ABSTRACT = {Crowd counting is a promising hotspot of computer vision involving crowd intelligence analysis, achieving tremendous success recently with the development of deep learning. However, there have been still many challenges including crowd multi-scale variations and high network complexity, etc. To tackle these issues, a lightweight Res-connection multi-branch network (LRMBNet) for highly accurate crowd counting and localization is proposed. Specifically, using improved ShuffleNet V2 as the backbone, a lightweight shallow extractor has been designed by employing the channel compression mechanism to reduce enormously the number of network parameters. A light multi-branch structure with different expansion rate convolutions is demonstrated to extract multi-scale features and enlarged receptive fields, where the information transmission and fusion of diverse scale features is enhanced via residual concatenation. In addition, a compound loss function is introduced for training the method to improve global context information correlation. The proposed method is evaluated on the SHHA, SHHB, UCF-QNRF and UCF_CC_50 public datasets. The accuracy is better than those of many advanced approaches, while the number of parameters is smaller. The experimental results show that the proposed method achieves a good tradeoff between the complexity and accuracy of crowd counting, indicating a lightweight and high-precision method for crowd counting.},
DOI = {10.32604/cmc.2024.048928}
}