Involution改进的卷积神经网络人群计数方法  

Convolutional Neural Network Method for Crowd Counting Improved using Involution Operator

作　　者：李兆鑫 卢树华[1] 兰凌强 刘淇缘 Li Zhaoxin;Lu Shuhua;Lan Lingqiang;Liu Qiyuan(College of Information and Cyber Security,People’s Public Security University of China,Beijing 102600,China)

机构地区：[1]中国人民公安大学信息网络安全学院,北京102600

出　　处：《激光与光电子学进展》2022年第18期251-258,共8页Laser & Optoelectronics Progress

基　　金：公安学科基础理论研究专项(2021XKZX08);中央高校基本科研业务经费重大项目(2021JKF102)。

摘　　要：针对现有人群计数方法大多采用卷积操作提取特征,空间多样性特征信息提取和传递能力不足的问题,提出一种Involution改进的单列深层人群计数网络。该网络以VGG-16为基本框架,引入Involution算子替代卷积操作,并辅以残差链接提高对空间特征信息的感知和传递能力;采用膨胀卷积保持分辨率的同时扩大感受野,丰富深度语义特征;利用联合损失函数监督网络训练,提高计数准确性和全局信息相关性。所提方法在公开数据集ShangHaiTech、UCF-QNRF、UCF_CC_50上的性能较基线模型提升显著,并超越了诸多当前的先进算法。结果表明:所提人群计数方法具有较高的准确性和更好的鲁棒性。Most existing crowd counting methods use convolution operations to extract features.However,extracting and transmitting spatial diversity feature information are difficult.In this paper,we propose an Involutionimproved singlecolumn deep crowdcounting network to mitigate these problems.Using VGG16 as the backbone,the proposed network uses an Involution operator combined with residual connection to replace the convolution operation,thereby enhancing the perception and transmission for spatial feature information.The dilated convolution was adopted to expand the receptive field while maintaining resolution to enrich deep semantic features.Additionally,we used the joint loss function to supervise the network training,improving counting accuracy and global information correlation.Compared with the baseline model,the performance of the proposed method across the ShangHaiTech,UCFQNRF,and UCF_CC_50 datasets considerably is improved,demonstrating that our approach outperforms many current advanced algorithms.Furthermore,results show that the proposed crowd counting method has higher accuracy and better robustness than other methods.

关 键 词：人群计数 Involution算子 膨胀卷积 全局损失 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]