首页 | 本学科首页   官方微博 | 高级检索  
     

卷积自编码器中粗粒度池化特征提取研究
引用本文:罗畅,王洁,王鹏飞,肖军,肖红. 卷积自编码器中粗粒度池化特征提取研究[J]. 电子学报, 2017, 45(10): 2390-2401. DOI: 10.3969/j.issn.0372-2112.2017.10.012
作者姓名:罗畅  王洁  王鹏飞  肖军  肖红
作者单位:1. 空军工程大学防空反导学院, 陕西西安 710051;2. 94691部队, 福建连城 366202
基金项目:国家自然科学基金(71501184)
摘    要:卷积自编码器(Convolutional Auto Encoder,CAE)提取的粗粒度池化特征具有一定范围内旋转和平移的不变性,因而得到广泛使用.然而,目前CAE仍主要依靠经验调节内部参数以获取满足要求的粗粒度池化特征.本文将CAE看作一个整体,从概率上分析了影响其表现的具体原因,构建了一个通用框架用于调节其中的主要参数以获取更好的粗粒度特征.首先从概率上权衡了粗粒度特征在池化层上的判别性与不变性,并在CAE中选择合适的卷积范围和白化参数.然后通过分析池化域内特征的稀疏度选择相应的池化方法以获取具有更好可分离性的粗粒度池化特征.在两个公开数据库(STL-10和CIFAR-10)的实验结果表明本文提出的方法可以指导CAE提取到更好的粗粒度池化特征并在多类分类任务中表现得更好.

关 键 词:粗粒度特征  池化  卷积自编码器  非监督学习  深度学习  
收稿时间:2016-03-29

Coarse-Grained Pooled Features Learning in Convolutional Autoencoders
LUO Chang,WANG Jie,WANG Peng-fei,XIAO Jun,XIAO Hong. Coarse-Grained Pooled Features Learning in Convolutional Autoencoders[J]. Acta Electronica Sinica, 2017, 45(10): 2390-2401. DOI: 10.3969/j.issn.0372-2112.2017.10.012
Authors:LUO Chang  WANG Jie  WANG Peng-fei  XIAO Jun  XIAO Hong
Affiliation:1. Air and Missile Defense College, Air Force Engineering University, Xi'an, Shaanxi 710051, China;2. Unit 94691 PLA, Liancheng, Fujian 366202, China
Abstract:Coarse-grained pooled features obtained from convolutional autoencoder (CAE) achieve scale and shift invariances and have been widely used recently.However,in most previous works coarse-grained pooled features are obtained by empirically modulating parameters in CAE.In this paper,we see the CAE as a whole,find the probabilistic factors affecting the performance of it,and formulate a general framework to regulate parameters in it to obtain better coarse-grained representation.Firstly,the discrimination-invariance tradeoff of coarse-grained features is probabilistically evaluated in the pooled feature maps.Furthermore,the proper convolved filter scales and appropriate whitening parameters are suggested in a CAE.Secondly,pooling approaches are combined with the sparsity degree in pooling regions,and we propose the preferable pooling approach in different cases.Experimental results on two independent benchmark datasets (STL-10 and CIFAR-10) demonstrate that our framework can guide CAEs to extract better coarse-grained pooled features and performs better in multiclass classification task.
Keywords:coarse-grained features  pooling  convolutional autoencoder  unsupervised learning  deep learning
本文献已被 万方数据 等数据库收录!
点击此处可从《电子学报》浏览原始摘要信息
点击此处可从《电子学报》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号