|
|
论文:2021,Vol:39,Issue(5):1122-1129 |
|
|
引用本文： |
|
|
王耀力, 刘晓慧, 李斌, 常青. 流形嵌入的选择性伪标记与小样本数据迁移[J]. 西北工业大学学报 |
|
|
WANG Yaoli, LIU Xiaohui, LI Bin, CHANG Qing. The manifold embedded selective pseudo-labeling algorithm and transfer learning of small sample dataset[J]. Northwestern polytechnical university |
|
|
|
|
|
|
|
| 流形嵌入的选择性伪标记与小样本数据迁移 |
|
| 王耀力1, 刘晓慧1, 李斌2, 常青1 |
|
1. 太原理工大学 信息与计算机学院, 山西 太原 030024;
2. 西北工业大学 航海学院, 陕西 西安 710071 |
| 摘要: |
| 特殊场景分类和识别任务面临样本不易获得而造成样本缺乏，利用源域（或称辅助域）数据构建领域自适应迁移学习模型，提高小样本机器学习在这些困难场景中的分类准确度与性能是当前研究的热点与难点。提出深度卷积与格拉斯曼流形嵌入的选择性伪标记算法（deep convolution and Grassmann manifold embedded selective pseudo-labeling，DC-GMESPL）模型，以实现在多种小样本数据集间迁移学习分类。针对目标域特殊场景，如森林火灾烟雾视频图像的本地样本数据缺乏情景，使用卫星遥感图像异地样本数据作为源域，基于Resnet50深度迁移网络，同时提取源域与目标域的烟雾特征；通过去除源域特征间的相关性，并与目标域重新关联，最小化源域与目标域特征分布距离，使源域与目标域特征分布对齐；在格拉斯曼流形空间中，用选择性伪标记算法对目标域数据作伪标记；构建一种可训练模型完成小样本数据间迁移分类。通过卫星遥感图像与视频影像数据集间迁移学习，对文中模型进行评估。实验表明，DC-GMESPL迁移准确率均高于DC-CMEDA、Easy TL、CMMS和SPL等方法。与作者先期研究的DC-CMEDA算法相比，新算法DC-GMESPL的准确率得到进一步提升；DC-GMESPL从卫星遥感图像到视频图像迁移准确率提高了0.50%，而从视频图像到卫星遥感图像迁移准确率提高了8.50%，且在性能上有了很大改善。 |
| 关键词:
迁移学习
领域自适应
深度卷积神经网络
小样本数据集
森林火灾烟雾特征
|
|
| The manifold embedded selective pseudo-labeling algorithm and transfer learning of small sample dataset |
|
| WANG Yaoli1, LIU Xiaohui1, LI Bin2, CHANG Qing1 |
|
1. College of Information and Computer, Taiyuan University of Technology, Taiyuan 030024, China;
2. School of Marine Science and Technology, Northwestern Polytechnical University, Xi'an 710072, China |
| Abstract: |
| Special scene classification and identification tasks are not easily fulfilled to obtain samples, which results in a shortage of samples. The focus of current researches lies in how to use source domain data (or auxiliary domain data) to build domain adaption transfer learning models and to improve the classification accuracy and performance of small sample machine learning in these special and difficult scenes. In this paper, a model of deep convolution and Grassmann manifold embedded selective pseudo-labeling algorithm (DC-GMESPL) is proposed to enable transfer learning classifications among multiple small sample datasets. Firstly, DC-GMESPL algorithm uses satellite remote sensing image sample data as the source domain to extract the smoke features simultaneously from both the source domain and the target domain based on the Resnet50 deep transfer network. This is done for such special scene of the target domain as the lack of local sample data for forest fire smoke video images. Secondly, DC-GMESPL algorithm makes the source domain feature distribution aligned with the target domain feature distribution. The distance between the source domain and the target domain feature distribution is minimized by removing the correlation between the source domain features and re-correlation with the target domain. And then the target domain data is pseudo-labeled by selective pseudo-labeling algorithm in Grassmann manifold space. Finally, a trainable model is constructed to complete the transfer classification between small sample datasets. The model of this paper is evaluated by transfer learning between satellite remote sensing image and video image datasets. Experiments show that DC-GMESPL transfer accuracy is higher than DC-CMEDA, Easy TL, CMMS and SPL respectively. Compared with our former DC-CMEDA, the transfer accuracy of our new DC-GMESPL algorithm has been further improved. The transfer accuracy of DC-GMESPL from satellite remote sensing image to video image has been improved by 0.50%, the transfer accuracy from video image to satellite remote sensing image has been improved by 8.50% and then, the performance has been greatly improved. |
| Key words:
transfer learning
domain adaptation
deep convolution neural networks
small sample dataset
forest fire smoke features
|
|
| 收稿日期: 2021-02-01
修回日期:
|
| DOI: 10.1051/jnwpu/20213951122 |
| 基金项目: 国家自然科学基金（61828601）与山西省重点研发项目（201903D321003）资助 |
| 通讯作者:
Email： |
| 作者简介: 王耀力(1965-),太原理工大学副教授、博士,主要从事计算智能、机器视觉研究。
|
|
| 相关功能 |
|
|
|
|
| 作者相关文章 |
|
| 王耀力 在本刊中的所有文章 |
| 刘晓慧 在本刊中的所有文章 |
| 李斌 在本刊中的所有文章 |
| 常青 在本刊中的所有文章 |
|
|
|
|
|
|
|
|
参考文献: |
|
|
[1] WANG Yaoli, WANG Lipo, YANG Fangjun, et al. Advantages of direct input-to-output connections in neural networks:the Elman network for stock index forecasting[J]. Information Sciences, 2021, 547:1066-1079
[2] WANG Yaoli, LIU Xiaohui, LI Maozhen, et al. Deep convolution and correlated manifold embedded distribution alignment for forest fire smoke prediction[J]. Computing and Informatics, 2020, 39(1/2):318-339
[3] KARLWEISS T K, WANG D. A survey of transfer learning[J]. Journal of Big Data, 2016, 3(9):1-40
[4] FERNANDO B, HABRARD A, SEBBAN M, et al. Unsupervised visual domain adaptation using subspace alignment[C]//Proceedings of the IEEE International Conference on Computer Vision, 2013:2960-2967
[5] SUN Baochen, SAENKO Kate. Subspace distribution alignment for unsupervised domain[C]//Proceedings of BMVC, 2015
[6] SUN Baochen, FENG Jiashi, SAENKO Kate. Return of frustratingly dasy domain adaptation[C]//Proceedings of 30th AAAI Conference on Artificial Intelligence, 2016
[7] GHIFARY M, BALDUZZI D, KLEIJN W B, et al. Scater component analysis:a unified framework for domain adaptation and domain generalization[J]. IEEE Trans on Pattern Analysis and Machine Intelligence, 2017, 39(7):1414-1430
[8] GOPALAN R, LI R, CHELLAPPA R. Domain adaptation for object recognition:an unsupervised approach[C]//Proceedings of the IEEE International Conference on Computer Vision, 2011
[9] GONG Boqing, SHI Yuan, SHA Fei, et al. Geodesic flow kernel for unsupervised domain adaptation[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2012
[10] BAKTASHMOTLAGH Mahsa, HARANDI Mehrtash, SALZMANN Mathieu. Distribution-matching embedding for visual domain adaptation[J]. Journal of Machine Learning Research, 2016, 17:3760-3789
[11] BAKTASHMOTLAGH M, HARANDI M T, LOVELL B C, et al. Domain adaptation on the statistical manifold[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2014
[12] GONG M, ZHANG K, LIU T, et al. Domain adaptation with conditional transferable components[C]//Proceedings of International Conference on Machine Learning, 2016
[13] PAN S J, TSANG I W, KWOK J T, et al. Domain adaptation via transfer component analysis[J]. IEEE Trans on Neural Networks, 2011, 22(2):199-210
[14] DORRI F, GHODSI A. Adapting component analysis[C]//Proceedings of the IEEE 12th International Conference on Data Mining, 2012:846-851
[15] DUAN L, TSANG I W, XU D. Domain transfer multiple kernel learning[J]. IEEE Trans on Pattern Analysis and Machine Intelligence, 2012, 34(3):465-479
[16] ZELLINGER W, GRUBINGER T, LUGHOFER E, et al. Central moment discrepancy (CMD) for domain-invariant representation learning[C]//Proceedings of the International Conference on Learning Representations, 2017
[17] LONG M, WANG J, DING G, et al. Transfer feature learning with joint distribution adaptation[C]//IEEE International Conference on Computer Vision, 2013
[18] WANG Jindong, CHEN Yiqiang, HAO Shuji, et al. Balanced distribution adaptation for transfer learning[C]//Proceedings of the IEEE International Conference on Data Mining, 2017:1129-1134
[19] WANG Jindong, FENG Wenjie, CHEN Yiqiang, et al. Visual domain adaptation with manifold embedded distribution alignment[C]//Proceedings of the 26th ACM International Conference on Multimedia, 2018
[20] BU Fengju, MOHAMMAD Samadi Gharajeh. Intelligent and vision-based fire detection systems:a survey[J]. Image and Vision Computing, 2019, 91(2):103803
[21] HE Xiaofei, NIYOGI Partha. Locality preserving projections[C]//Proceedings of the 16th International Conference on Neural Information Processing Systems, 2003:153-160
[22] PEI Zhongyi, CAO Zhangjie, LONG Mingsheng, et al. Multi-adversarial domain adaptation[C]//Proceedings of the AAAI Conference on Artificial Intelligence, 2018 |
|
|
|
|
|
|
|
