Delving Deep into Multiscale Pedestrian Detection via Single Scale Feature Maps

Sensors

卷:18

Jie Wu¹  Weinan Zhang²  Rui Yu³  Shihai Shao⁴  Xinchuan Fu⁵ 

[1] Engineering, Shanghai Jiao Tong University, Shanghai 200240, China;

[2] Department of Computer Science &

[3] Department of Computer Science, University College London, London WC1E 6BT, UK;

[4] Department of MOE Research Center for Software/Hardware Co-Design Engineering and Application, East China Normal University, Shanghai 200062, China;

[5] National Key Laboratory of Science and Technology on Communications, University of Electronic Science and Technology of China, Chengdu 611731, China;

关键词: pedestrian detection;    boosted decision tree;    scale invariance;    receptive field correspondence;    soft decision tree;   

DOI  :  10.3390/s18041063

来源: DOAJ

【 摘 要 】

The standard pipeline in pedestrian detection is sliding a pedestrian model on an image feature pyramid to detect pedestrians of different scales. In this pipeline, feature pyramid construction is time consuming and becomes the bottleneck for fast detection. Recently, a method called multiresolution filtered channels (MRFC) was proposed which only used single scale feature maps to achieve fast detection. However, there are two shortcomings in MRFC which limit its accuracy. One is that the receptive field correspondence in different scales is weak. Another is that the features used are not scale invariance. In this paper, two solutions are proposed to tackle with the two shortcomings respectively. Specifically, scale-aware pooling is proposed to make a better receptive field correspondence, and soft decision tree is proposed to relive scale variance problem. When coupled with efficient sliding window classification strategy, our detector achieves fast detecting speed at the same time with state-of-the-art accuracy.

【 授权许可】

Unknown