AUTHOR=Yun Juntong , Jiang Du , Liu Ying , Sun Ying , Tao Bo , Kong Jianyi , Tian Jinrong , Tong Xiliang , Xu Manman , Fang Zifan 

TITLE=Real-Time Target Detection Method Based on Lightweight Convolutional Neural Network

JOURNAL=Frontiers in Bioengineering and Biotechnology

VOLUME=Volume 10 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/bioengineering-and-biotechnology/articles/10.3389/fbioe.2022.861286

DOI=10.3389/fbioe.2022.861286

ISSN=2296-4185

ABSTRACT=The continuous development of deep learning makes the target detection technology mature day by day. The current research focuses on improving the accuracy of the target detection technology, resulting in the target detection model is too large. The number of parameters and detection speed of target detection model are very important for the practical application of target detection technology in embedded system, which is also an urgent problem to be solved. This paper proposed a real-time target detection method based on lightweight convolutional neural network to reduce the number of model parameters and improve the detection speed. In this paper, the depth-detachable residual module is constructed by combining depth-detachable convolution and bottleneck free residual module, and the depth-detachable residual module and depth-detachable convolution structure are used to replace the VGG backbone network in SSD network for feature extraction of target detection model. This paper used the self-built target detection data set in complex scenes for comparative experiments, the experimental results verify the effectiveness and superiority of the proposed method. The model was tested on video to verify the real-time performance of the model, and the model was deployed on Android platform to verify the scalability of the model.