期刊
EXPERT SYSTEMS WITH APPLICATIONS
卷 197, 期 -, 页码 -出版社
PERGAMON-ELSEVIER SCIENCE LTD
DOI: 10.1016/j.eswa.2022.116662
关键词
Crowd counting; Lightweight convolutional neural network; Scale variability; Feature fusion; Scale adaptation
类别
资金
- National Natural Science Foundation of China [61972134]
- Henan Provincial Department of Science and Technology Research Project [192102210118]
This study proposes a Lightweight Multi-Scale Adaptive Network (LigMSANet) to address the challenges of scale variation and real-time counting in highly congested scenes. The method breaks the scale limitation and adjusts the proportion of neurons with different receptive field sizes through a novel multi-scale adaptation module. By replacing standard convolution with depthwise separable convolution and using a tailored MobileNetV2, the model achieves improved performance with fewer parameters and runtimes.
Scale variation and real-time counting are challenging problems for crowd counting in highly congested scenes. To remedy these issues, we proposed a Lightweight Multi-Scale Adaptive Network (LigMSANet). There are two strong points in our method. First, the scale limitation is broken and the proportion of neurons with different receptive field sizes are adjusted spontaneously according to input images through a novel multi-scale adaptation module (MSAM). Second, the model performance is significantly improved at a little cost of parameter by replacing the standard convolution with the depthwise separable convolution and a tailored MobileNetV2 with 5 bottleneck blocks (here, the step size of the fourth bottleneck block is 1). To demonstrate the effectiveness of the proposed method, we conduct extensive experiments on three major crowd counting datasets (ShanghaiTech, UCF_CC_50 and UCSD) and our method achieves superior performance to state-of-the-art methods while with much less parameters and runtimes.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据