[1]杨 虹,范 勇.一种基于区分区域定位的细粒度图像识别方法[J].计算机技术与发展,2023,33(11):169-174.[doi:10. 3969 / j. issn. 1673-629X. 2023. 11. 025]
　YANG Hong,FAN Yong.A Fine Grain Image Recognition Method Based on Distinguishable Region Location[J].,2023,33(11):169-174.[doi:10. 3969 / j. issn. 1673-629X. 2023. 11. 025]
点击复制

一种基于区分区域定位的细粒度图像识别方法()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
33
期数:
2023年11期
页码:
169-174
栏目:
人工智能
出版日期:
2023-11-10

文章信息/Info

Title:
A Fine Grain Image Recognition Method Based on Distinguishable Region Location
文章编号:
1673-629X(2023)11-0169-06
作者:
杨 虹; 范 勇
西南科技大学 计算机科学与技术学院,四川 绵阳 621010
Author(s):
YANG Hong; FAN Yong
School of Computer Science and Technology,Southwest University of Science and Technology,Mianyang 621010,China
关键词:
细粒度图像识别; 通道注意力; 标签平滑; 区域定位; 特征提取
Keywords:
fine-grained image recognition; channel attention; label smoothing; region location; feature extraction
分类号:
TP391. 4
DOI:
10. 3969 / j. issn. 1673-629X. 2023. 11. 025
摘要:
细粒度图像识别的目标为区分大类对象中的子类对象,由于子类对象间差别细微,使得细粒度图像识别较为困难。 为此,提出一种基于区分区域定位的细粒度图像识别方法。 首先由贝叶斯个性化排序损失( Bayesian PersonalizedRanking Loss,BPRLoss) 监督区域提议网络提议一些重要的局部区域,随后采用引入高效通道注意力模块的特征提取器提取局部区域的细粒度特征进行识别。 同时采用标签平滑策略使同类靠近,不同类远离以监督网络学习对象有区别的特征,进一步促进网络定位区分区域。 实验结果表明,所提方法在三种通用的细粒度图像识别数据集 CUB - 200 - 2011、FGVC Aircraft、Stanford Cars 上取得了较高的识别准确率,分别为 89. 0% 、93. 9% 、94. 3% ,相比导航网络( NTS-Net) 有显著提升,分别提升 1. 5 百分点、2. 5 百分点和 0. 4 百分点。 同时,所提方法较 NTS-Net 能够更为有效地定位区分区域和提取图像的细粒度特征。
Abstract:
The goal of fine- grained image recognition is to distinguish sub - class objects in large class objects. Because of the subtledifferences between sub-class objects,fine-grained image recognition is more difficult. For this reason,a fine-grained image recognitionmethod based on differentiated region location is proposed. Firstly, the Bayesian Personalized Ranking Loss ( BPRLoss) supervisedregion proposes that the network proposes some important local regions, and then uses the feature extractor introducing the efficientchannel attention module to extract the fine-grained features of the local regions for recognition. At the same time,the tag smoothingstrategy is used to make the same class close and different classes far away to monitor the different characteristics of the network learningobjects,and further promote the network location to distinguish regions. The experimental results show that the proposed method hasachieved high recognition accuracy on three common fine - grained image recognition data sets CUB - 200 - 2011, FGVC Aircraft andStanford Cars,which are 89. 0% , 93. 9% and 94. 3% , respectively. Compared with the navigation network ( NTS - Net ) , it hassignificantly improved by 1. 5 percentage points,2. 5 percentage points and 0. 4 percentage points respectively. At the same time,theproposed method is more effective than NTS-Net in locating and distinguishing regions and extracting fine-grained features of images.

相似文献/References:

[1]姜孟超,范灵毓,李硕豪*.基于注意力双线性池化的细粒度舰船识别[J].计算机技术与发展,2022,32(08):66.[doi:10. 3969 / j. issn. 1673-629X. 2022. 08. 011]
　JIANG Meng-chao,FAN Ling-yu,LI Shuo-hao*.Weakly Supervised Fine-grained Natural Scene Ship Recognition viaAttention Bilinear Pooling[J].,2022,32(11):66.[doi:10. 3969 / j. issn. 1673-629X. 2022. 08. 011]
[2]张 伟,刘宁钟,寇金桥.基于深度特征金字塔的路面病害检测[J].计算机技术与发展,2022,32(12):173.[doi:10. 3969 / j. issn. 1673-629X. 2022. 12. 026]
　ZHANG Wei,LIU Ning-zhong,KOU Jin-qiao.Pavement Disease Detection Based on Depth Feature Pyramids[J].,2022,32(11):173.[doi:10. 3969 / j. issn. 1673-629X. 2022. 12. 026]
[3]谢紫薇,鲁大营*,李志琦,等.基于扩张卷积与注意力的甲状腺超声分割方法[J].计算机技术与发展,2023,33(03):71.[doi:10. 3969 / j. issn. 1673-629X. 2023. 03. 011]
　XIE Zi-wei,LU Da-ying*,LI Zhi-qi,et al.Dilated Convolution and Attention-based Ultrasound Segmentation of Thyroid Nodules[J].,2023,33(11):71.[doi:10. 3969 / j. issn. 1673-629X. 2023. 03. 011]
[4]关 慧,曹同洲.基于 CNN 和多注意力机制的 XSS 检测模型[J].计算机技术与发展,2023,33(04):175.[doi:10. 3969 / j. issn. 1673-629X. 2023. 04. 026]
　GUAN Hui,CAO Tong-zhou.XSS Detection Model Based on CNN and Multi-attention Mechanism[J].,2023,33(11):175.[doi:10. 3969 / j. issn. 1673-629X. 2023. 04. 026]
[5]周 帅,李 理,彭章君,等.基于多通道特征和混合注意力的环境声音分类[J].计算机技术与发展,2023,33(08):43.[doi:10. 3969 / j. issn. 1673-629X. 2023. 08. 007]
　ZHOU Shuai,LI Li,PENG Zhang-jun,et al.Environmental Sound Classification Based on Multi-channel Features and Mixed Attention[J].,2023,33(11):43.[doi:10. 3969 / j. issn. 1673-629X. 2023. 08. 007]
[6]宁园园,张素兰,陈 飞.基于双注意力机制的零样本建筑图像分类方法[J].计算机技术与发展,2023,33(10):35.[doi:10. 3969 / j. issn. 1673-629X. 2023. 10. 006]
　NING Yuan-yuan,ZHANG Su-lan,CHEN Fei.Zero-shot Architectural Image Classification Method Based on Dual Attention Mechanism[J].,2023,33(11):35.[doi:10. 3969 / j. issn. 1673-629X. 2023. 10. 006]
[7]姚亮亮,张太红*,张洋宁,等.单参数通道注意力模块[J].计算机技术与发展,2023,33(12):215.[doi:10. 3969 / j. issn. 1673-629X. 2023. 12. 030]
　YAO Liang-liang,ZHANG Tai-hong*,ZHANG Yang-ning,et al.Single-parameter Channel Attention Module[J].,2023,33(11):215.[doi:10. 3969 / j. issn. 1673-629X. 2023. 12. 030]

更新日期/Last Update: 2023-11-10