基于跨领域主动学习的图像分类方法

doi:10.11772/j.issn.1001-9081.2014.04.1169

计算机应用 ›› 2014, Vol. 34 ›› Issue (4): 1169-1171.DOI: 10.11772/j.issn.1001-9081.2014.04.1169

基于跨领域主动学习的图像分类方法

邵忻

天津外国语大学基础课教学部,天津 300204

收稿日期:2013-10-15 修回日期:2013-12-17 出版日期:2014-04-01 发布日期:2014-04-29
通讯作者: 邵忻
作者简介:邵忻(1979-),女,天津人,讲师,硕士,主要研究方向:计算机软件。

Cross-domain active learning algorithm for image classification

SHAO Xin

Faculty of Fundamental Courses, Tianjin Foreign Studies University, Tianjin 300204, China

Received:2013-10-15 Revised:2013-12-17 Online:2014-04-01 Published:2014-04-29
Contact: SHAO Xin

摘要/Abstract

摘要：

针对基于单一领域主动学习的图像分类方法不能利用不同领域图像共同特征导致标记效率低下的问题，提出一种基于跨领域主动学习的图像分类方法。由不同领域图像学习出含共同隐特征的子空间，综合考虑共同特征和领域相关特征，将数据实例引起的模型损失减少量分解到一个共同部分和领域相关部分，从而领域间的共同信息可以编码到模型损失减少的共同部分并用来进行查询。实验结果显示该方法相对于单一模型学习和混合模型学习方法可以减少将近30%的标记工作，并且可以获得更高的精度，表明该方法可以更高效地运用于各种图像分类任务。

Abstract:

To solve the problem of ignoring common information in different domains in traditional active learning image classification, a new multi-domain active learning image classification method was proposed to reduce the labeling effort of image classification. It learned a subspace to represent common features among different images. Considering the common features and domain-specific features, the model loss due to each data instance could be divided into two parts, so that the common information could be queried from the common part. The experimental results show that the new method has some precise increase and has 30% less labeling efforts than the single model method and mixture model method. The results reveal that the new method can be widely used in all kinds of image classification tasks with higher precise and efficiency.

中图分类号:

TP391.41

邵忻. 基于跨领域主动学习的图像分类方法[J]. 计算机应用, 2014, 34(4): 1169-1171.

SHAO Xin. Cross-domain active learning algorithm for image classification[J]. Journal of Computer Applications, 2014, 34(4): 1169-1171.

参考文献

［1］WANG J, YANG J, YU K, et al.Learning locality-constrained linear coding for image classification ［C］// CVPR2010: Proceedings of the 2010 IEEE Conference on Computer Vision and Pattern Recognition. Washington, DC: IEEE Computer Society, 2010:3360-3367.
［2］LIU K, QIAN X, WANG Z. Survey on active learning algorithms ［J］. Computer Engineering and Applications, 2012, 48(34): 1-6. (刘康,钱旭,王自强.主动学习算法研究综述［J］.计算机工程与应用,2012,48 (34):1-6.)
［3］HOI S C H, JIN R, LYU M R. Batch mode active learning with applications to text categorization and image retrieval ［J］. IEEE Transactions on Knowledge and Data Engineering, 2009, 21(9): 1233-1248.
［4］CHEN R, CAO Y. Multi-class image classification with active learning and semi-supervised learning ［J］. Acta Automatica Sinica, 2011, 37(8): 954-962. (陈荣,曹永锋,孙洪.基于主动学习和半监督学习的多类图像分类［J］.自动化学报,2011,37(8):954-962.)
［5］MAJI S, BERGY A, MALIK J. Classification using intersection kernel support vector machines is efficient ［C］// CVPR2008: Proceedings of the 2008 IEEE Conference on Computer Vision and Pattern Recognition. Washington, DC: IEEE Computer Society, 2008: 1-8.
［6］CHEN J, YE Z. Remote sensing image classification using active learning with multiclass support vector machine ［J］. Science of Surveying and Mapping, 2009, 34(4): 34-39. (陈建杰,叶智宣.多分类SVM主动学习及其在遥感图像分类中的应用［J］.测绘科学,2009, 34(4):34-39.)
［7］PAN S, NI X, SUN J, et al.Cross-domain sentiment classification via spectral feature alignment ［C］// WWW2010: Proceedings of the 19th International Conference on World Wide Web. New York: ACM, 2010: 751-760.
［8］TONG S, KOLLER D. Support vector machine active learning with applications to text classification ［J］. Journal of Mechanism Learning, 2002, 14(2): 45-66.
［9］DAGAN I, ENGELSON S P. Committee-based sampling for training probabilistic classifiers ［C］// ICML1995: Proceedings of the 12th International Conference on Machine Learning. San Francisco: Morgan Kaufmann Publishers Inc., 1995: 150-157.
［10］LEWIS D D, GALE W A. A sequential algorithm for training text classifiers ［C］// SIGIR'94: Proceedings of the 17th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, LNCS 1513. Berlin: Springer-Verlag, 1994: 3-12.
［11］TUIA D, RATLE F, PACIFICI F, et al.Active learning of very-high resolution optical imagery with SVM: entropy vs margin sampling ［C］// IGARSS 2008: Proceedings of the 2008 IEEE International Conference on Geoscience and Remote Sensing Symposium. Washington, DC: IEEE Computer Society, 2008, 4: 73-76.
［12］ASUNCION A, NEWMAN D J. UCI machine learning repository［EB/OL］. (2010-01-10) ［2013-01-15］.
http://archive.ics.uci.edu/ml/datasets.html.

[1]	章悦, 张亮, 谢非, 杨嘉乐, 张瑞, 刘益剑. 基于实例分割模型优化的道路抛洒物检测算法[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3228-3233.
[2]	孙琳, 袁玉波. 基于人眼状态的瞌睡识别算法[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3213-3218.
[3]	葛晨宇, 董良, 许伊昆, 常毅, 张宏鸣. 基于总变分低秩组稀疏的全球雷达数据修复算法[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3353-3361.
[4]	闫钧华, 侯平, 张寅, 吕向阳, 马越, 王高飞. 基于多尺度多分类器卷积神经网络的混合失真类型判定方法[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3178-3184.
[5]	李福海, 蒋慕蓉, 杨磊, 谌俊毅. 基于生成对抗网络的梯度引导太阳斑点图像去模糊方法[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3345-3352.
[6]	曹建芳, 闫敏敏, 贾一鸣, 田晓东. 融合迁移学习的Inception-v3模型在古壁画朝代识别中的应用[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3219-3227.
[7]	谌贵辉, 刘会康, 李忠兵, 彭娇, 汪少天, 林瑾瑜. 基于仲裁机制的生成对抗网络改进算法[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3185-3191.
[8]	杨鼎康黄帅王顺利翟鹏李一丹张立华. 基于对抗生成网络和网络集成的面部表情识别方法EE-GAN[J]. 计算机应用, 0, (): 0-0.
[9]	秦庭威赵鹏程秦品乐曾建朝柴锐黄永琦. 基于残差注意力机制的点云配准算法[J]. 计算机应用, 0, (): 0-0.
[10]	魏淳武, 赵涓涓, 唐笑先, 强彦. 基于多时期蒸馏网络的随访数据知识提取方法[J]. 计算机应用, 2021, 41(10): 2871-2878.
[11]	李自强, 王正勇, 陈洪刚, 李林怡, 何小海. 基于外观和动作特征双预测模型的视频异常行为检测[J]. 计算机应用, 2021, 41(10): 2997-3003.
[12]	裴仪瑶, 郭会明, 张丹普, 陈文博. 基于定位不确定性的鲁棒3D目标检测方法[J]. 计算机应用, 2021, 41(10): 2979-2984.
[13]	许学斌, 张佳达, 刘伟, 路龙宾, 赵雨晴. 融合空间和通道特征的高精度乳腺癌分类方法[J]. 计算机应用, 2021, 41(10): 3025-3032.
[14]	汪虹余, 张彧, 杨恒, 穆楠. 基于蚁群优化算法的弱光图像显著性目标检测[J]. 计算机应用, 2021, 41(10): 2970-2978.
[15]	张怡孙永荣赵科东李华曾庆化. 空中加油场景下的目标联合检测跟踪算法[J]. 计算机应用, 0, (): 0-0.

基于跨领域主动学习的图像分类方法

Cross-domain active learning algorithm for image classification

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics