Research Article
Pedestrian Detection Based on Deep Learning Under the Background of University Epidemic Prevention
@INPROCEEDINGS{10.1007/978-3-030-98002-3_14, author={Ruiyan Du and Jia Zhao and Jiangfan Xie and Tian Wen}, title={Pedestrian Detection Based on Deep Learning Under the Background of University Epidemic Prevention}, proceedings={Cognitive Radio Oriented Wireless Networks and Wireless Internet. 16th EAI International Conference, CROWNCOM 2021, Virtual Event, December 11, 2021, and 14th EAI International Conference, WiCON 2021, Virtual Event, November 9, 2021, Proceedings}, proceedings_a={CROWNCOM \& WICON}, year={2022}, month={3}, keywords={YOLOv3 TensorFlow Pedestrian detection Pedestrian tracking Pedestrian counting}, doi={10.1007/978-3-030-98002-3_14} }- Ruiyan Du
Jia Zhao
Jiangfan Xie
Tian Wen
Year: 2022
Pedestrian Detection Based on Deep Learning Under the Background of University Epidemic Prevention
CROWNCOM & WICON
Springer
DOI: 10.1007/978-3-030-98002-3_14
Abstract
In the context of the current normalization of epidemic prevention, the nucleic acid detection process in colleges and universities is limited in human and material resources. Teachers and students who perform nucleic acid detection often cannot maintain a distance of more than one meter from others, and there is a pedestrian group behavior that has a large cross-infection safety hazard. This article uses Depthwise Separable Convolution to improve the YOLOv3 algorithm, and the improved network structure constructs a pedestrian detection, pedestrian tracking, pedestrian counting and pedestrian cluster system based on Deep Learning under the TensorFlow framework. The training parameters and training time of the improved network model are reduced to a certain extent, improved the operation efficiency of the network model. The advantage is that it realizes the function of monitoring centralized nucleic acid detection scenes in colleges and universities and assisting volunteers to maintain a reasonable order, which can effectively prevent cross-infection problems caused by cluster effects.
