文献类型：期刊文献

英文题名：ResNet based on feature-inspired gating strategy

作者：Miao, Jun[1];Xu, Shaowu[1];Zou, Baixian[2];Qiao, Yuanhua[3]

第一作者：Miao, Jun

通讯作者：Miao, J[1]

机构：[1]Beijing Informat Sci & Technol Univ, Beijing Key Lab Internet Culture & Digital Dissem, Beijing 100101, Peoples R China;[2]Beijing Union Univ, Coll Appl Arts & Sci, Beijing 100191, Peoples R China;[3]Beijing Univ Technol, Coll Appl Sci, Beijing 100124, Peoples R China

第一机构：Beijing Informat Sci & Technol Univ, Beijing Key Lab Internet Culture & Digital Dissem, Beijing 100101, Peoples R China

通讯机构：[1]corresponding author), Beijing Informat Sci & Technol Univ, Beijing Key Lab Internet Culture & Digital Dissem, Beijing 100101, Peoples R China.

年份：2022

卷号：81

期号：14

起止页码：19283-19300

外文期刊名：MULTIMEDIA TOOLS AND APPLICATIONS

收录：;EI(收录号：20211310149139);Scopus(收录号：2-s2.0-85103187432);WOS:【SCI-EXPANDED(收录号:WOS:000632798100001)】；

基金：This research is partially sponsored by Beijing Natural Science Foundation (No. 4202025), Beijing Municipal Education Commission Projects (Nos. KM201911232003 and KZ201910005008), and the Research Fund from Beijing Innovation Center for Future Chips (No. KYJJ2018004).

语种：英文

外文关键词：ResNet; Gating network; Gating residual network; Feature-inspired gating

摘要：CNN(Convolutional Neural Networks) is a hot topic in the field of pattern recognition., especially in the field of image recognition. And ResNet(Residual Networks) is a special kind of CNN. Compared with the general CNN structure, ResNet introduces the residual unit with an identity mapping. Identity mapping allows the deep layers to directly learn the data received by the shallow layers, which reduces the difficulty of network convergence to a certain extent. As a result, ResNet has a better learning ability, has achieved good performance in various types of image recognition work. The essence of the residual network is to fuse two types of features from different receptive fields, using the fused features instead of the output features of the previous layer as the learning object. But the implementation of feature fusion in original ResNet is adding the two features with equal weights. And this method ignores the fact that the contribution of features from different levels to the learning of the network may not be the same. In this paper, we introduce a feature-inspired gating strategy in the residual unit of ResNet, which allows the network giving different weights to different features, so that the implementation of the feature fusion can be transformed from adding features with equal weights into weighted summation with different weights. And through experiments, we proved that ResNet with gating strategy proposed in this paper can obtain higher recognition accuracy than original ResNet.