帮助 关于我们

返回检索结果

基于样本对元学习的小样本图像分类方法
A Few-Shot Image Classification Method by Pairwise-Based Meta Learning

查看参考文献30篇

文摘 本文针对小样本图像分类问题,提出一种基于样本对的元学习(Pairwise-based Meta Learning,PML)方法.利用传递迁移学习对预训练好的Resnet50模型进行微调,得到一个更适应小样本任务的特征编码器,将该特征编码器作为元学习模型的初始特征编码器来训练模型,进一步增强了元学习模型的泛化能力;同时,本文还基于支持集与查询集样本之间的相似性提出元损失函数(Meta Loss,ML),其考虑了特征空间中查询集所有样本的相互关系,以此来缩小正样本类内距离,增加正负样本类间距离,从而提高分类精度.实验结果表明,本文的方法在1-shot、5-shot任务上分别达到了77.65%、89.65%的分类精度,较最新的元学习方法Meta-baseline分别提高7.38%、5.65%.
其他语种文摘 In this paper, a pairwise-based meta learning(PML) method is proposed for few-shot image classification. Transitive transfer learning is used to fine tune the pre-trained Resnet50 model to get a feature encoder that is more suitable for few shot task. The feature encoder is used as the initial feature encoder of the meta-learning model to train the model, which further enhances the generalization ability of the meta-learning model. Based on the similarity between the support set and the query set samples, a meta loss(ML) function is proposed, which considers the relationship between all the samples of the query set in the feature space, so as to reduce the within-class distance of positive samples and increase the between-class distance of positive and negative samples, thus improving the classification accuracy.The experimental results show that the classification accuracy of the methods in this paper is 77.65% and 89.65% on 1-shot and 5-shot tasks, respectively, and it is 7.38% and 5.65% higher than the latest meta-learning method, Meta-baseline.
来源 电子学报 ,2022,50(2):295-304 【核心库】
DOI 10.12263/DZXB.20210453
关键词 小样本图像 ; 传递迁移学习 ; 元学习 ; 元损失函数
地址

武汉科技大学, 冶金自动化与检测技术教育部工程研究中心, 湖北, 武汉, 430081

语种 中文
文献类型 研究性论文
ISSN 0372-2112
学科 自动化技术、计算机技术
基金 国家自然科学基金面上项目 ;  湖北省重点研发计划
文献收藏号 CSCD:7195140

参考文献 共 30 共2页

1.  Vinyals O. Matching networks for one shot learning. Proceedings of the 30th International Conference on Neural Information Processing Systems,2016:3630-3638 CSCD被引 10 次    
2.  季鼎承. 基于域与样例平衡的多源迁移学习方法. 电子学报,2019,47(3):692-699 CSCD被引 8 次    
3.  Pan S J. A survey on transfer learning. IEEE Transactions on Knowledge & Data Engineering,2010,22(10):1345-1359 CSCD被引 1054 次    
4.  李凡长. 元学习研究综述. 计算机学报,2021,44(2):422-446 CSCD被引 51 次    
5.  Lake B. One shot learning of simple visual concepts. Proceedings of the Annual Meeting of the Cognitive Science Society. 33(33),2011 CSCD被引 1 次    
6.  Snell J. Prototypical networks for few-shot learning. Proceedings of the International Conference on Neural Information Processing Systems,2017:4077-4087 CSCD被引 1 次    
7.  Tan B. Transitive transfer learning. Proceedings of the 21st ACM SIGKDD International Conference on Knowledge Discovery and Data Mining,2015:1155-1164 CSCD被引 4 次    
8.  Sun Q. Meta-transfer learning for few-shot learning. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition,2019:403-412 CSCD被引 11 次    
9.  Chen Y. A new meta-baseline for few-shot learning. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition,2020:4390 CSCD被引 1 次    
10.  Lubbers N. Inferring low-dimensional microstructure representations using convolutional neural networks. Physical Review E,2017,96(5):052111 CSCD被引 3 次    
11.  Decost B L. Exploring the microstructure manifold: Image texture representations applied to ultrahigh carbon steel microstructures. Acta Materialia,2017,133:30-40 CSCD被引 12 次    
12.  Azimi S M. Advanced steel microstructural classification by deep learning methods. Scientific Reports,2018,8(1):2128 CSCD被引 28 次    
13.  Lee K. Meta-learning with differentiable convex optimization. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition,2019:10657-10665 CSCD被引 6 次    
14.  Ren M. Meta-learning for semi-supervised few-shot classification. Proceedings of the International Conference on Learning Representations CSCD被引 1 次    
15.  He K. Deep residual learning for image recognition. Proceedings of IEEE Conference on Computer Vision and Pattern Recognition,2016:770-778 CSCD被引 93 次    
16.  Wang X. Multi-similarity loss with general pair weighting for deep metric learning. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition,2019:5022-5030 CSCD被引 4 次    
17.  Weinberger K Q. Distance metric learning for large margin nearest neighbor classification. Journal of Machine Learning Research,2009,10(2):207-244 CSCD被引 102 次    
18.  Loshchilov I. Decoupled weight decay regularization. Proceedings of the European International Conference on Learning Representations(ICLR),2017 CSCD被引 1 次    
19.  Zeilerm D. Adaptive deconvolutional networks for mid and high level feature learning. Proceedings of the International Conference on Computer Vision,2011:2018-2025 CSCD被引 1 次    
20.  Finn C. Model-agnostic metalearning for fast adaptation of deep networks. Proceedings of the International Conference on Machine Learning,2017:1126-1135 CSCD被引 17 次    
引证文献 9 篇

1 李金灵 基于改进YOLOv5算法的带钢表面缺陷检测 钢铁研究学报,2023,35(6):767-777
CSCD被引 6 次

2 何玉林 基于数据生成模型的仿真样本点插补方法 系统仿真学报,2023,35(9):1948-1964
CSCD被引 0 次

显示所有9篇文献

论文科学数据集
PlumX Metrics
相关文献

 作者相关
 关键词相关
 参考文献相关

版权所有 ©2008 中国科学院文献情报中心 制作维护：中国科学院文献情报中心
地址：北京中关村北四环西路33号 邮政编码：100190 联系电话：(010)82627496 E-mail:cscd@mail.las.ac.cn 京ICP备05002861号-4 | 京公网安备11010802043238号