期刊
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY
卷 28, 期 8, 页码 1788-1797出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TCSVT.2016.2637379
关键词
Crowd counting; locality-aware feature (LAF); semantic attributes; weighted vector of locally aggregated descriptor (W-VLAD) encoder
资金
- National Natural Science Foundation of China [61473086, 61375001]
- open fund of the Key Laboratory of Measurement
- Control of Complex Systems of Engineering, Ministry of Education [MCCSE2013B01]
- NSF of Jiangsu Province [BK20140566, BK20150470]
- China Postdoctoral Science Foundation [2014M561586]
Crowd counting is an important task in computer vision, which has many applications in video surveillance. Although the regression-based framework has achieved great improvements for crowd counting, how to improve the discriminative power of image representation is still an open problem. Conventional holistic features used in crowd counting often fail to capture semantic attributes and spatial cues of the image. In this paper, we propose integrating semantic information into learning locality-aware feature (LAF) sets for accurate crowd counting. First, with the help of a convolutional neural network, the original pixel space is mapped onto a dense attribute feature map, where each dimension of the pixelwise feature indicates the probabilistic strength of a certain semantic class. Then, LAF built on the idea of spatial pyramids on neighboring patches is proposed to explore more spatial context and local information. Finally, the traditional vector of locally aggregated descriptor (VLAD) encoding method is extended to a more generalized form weighted-VLAD (W-VLAD) in which diverse coefficient weights are taken into consideration. Experimental results validate the effectiveness of our presented method.
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