A Prototype Learning Based Multi-Instance Convolutional Neural Network

被引：0
|
作者：
He K.-L. [1 ]
Shi Y.-H. [1 ]
Gao Y. [1 ]
Huo J. [1 ]
Wang D. [2 ]
Zhang Y. [2 ]
机构：
[1] State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing
[2] Bayi Hospital, Nanjing
来源：
Gao, Yang (gaoy@nju.edu.cn) | 2017年 / Science Press卷 / 40期
基金：
中国国家自然科学基金;
关键词：
Artificial intelligence; Convolutional neural network; Deep learning; Image classification; Multi-instance learning; Prototype learning;
D O I：
10.11897/SP.J.1016.2017.01265
中图分类号：
学科分类号：
摘要：
Convolutional neural network is a fully supervised deep learning model. It requires that the labels of samples are fully provided. In weakly supervised applications where labels of samples are partly provided, the usage of convolutional neural networks is greatly limited. To solve the weakly supervised multi-instance learning problem, a new multiple instance convolutional neural network is proposed. The proposed model introduces a new prototype learning layer into the network. The prototype learning layer uses a prototype based metric method to transform instance features into bag features. The network therefore can use label information of bag and learning the whole model in a compact process. The network is firstly tested on a lung cancer cell pathology image classification dataset. Results show, compared with hand designed image feature based methods, the proposed method achieved an improvement of about 12% in accuracy. Compared with convolutional neural network and multi-instance learning combined methods, the proposed method also achieved better results on all the evaluation criterion. Besides, the method is also tested on a natural image classification dataset (GRAZ-02). Comparable result is achieved by the proposed method compared with the state-of-the-art method. © 2017, Science Press. All right reserved.
引用
收藏
页码：1265 / 1274
页数：9
相关论文
共 29 条
  • [21] Cheng M.M., Zhang Z., Lin W.Y., Et al., BING: Binarized normed gradients for objectness estimation at 300 fps, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 3286-3293, (2014)
  • [22] Zhu L., Zhao B., Gao Y., Multi-class multi-instance learning for lung cancer image classification based on bag feature selection, Proceedings of the International Conference on Fuzzy Systems and Knowledge Discovery, 2, pp. 487-492, (2008)
  • [23] Chang C.C., Lin C.J., LIBSVM: A library for support vector machines, ACM Transactions on Intelligent Systems and Technology, 2, 3, (2011)
  • [24] Liu M., Zhang D., Shen D., Et al., Ensemble sparse classification of Alzheimer's disease, NeuroImage, 60, 2, pp. 1106-1116, (2012)
  • [25] Zhang L., Zhou W.D., Chang P.C., Et al., Kernel sparse representation-based classifier, IEEE Transactions on Signal Processing, 60, 4, pp. 1684-1695, (2012)
  • [26] Shi Y., Gao Y., Yang Y., Et al., Multimodal sparse representation-based classification for lung needle biopsy images, IEEE Transactions on Biomedical Engineering, 60, 10, pp. 2675-2685, (2013)
  • [27] Hegazy D., Denzler J., Generic object recognition using boosted combined features, Proceedings of the 2nd International Workshop on Robot Vision, pp. 355-366, (2008)
  • [28] Zhang Z., Li Z.N., Drew M.S., Learning image similarities via probabilistic feature matching, Proceedings of the IEEE International Conference on Image Processing, pp. 1857-1860, (2010)
  • [29] Gupta N., Das S., Chakraborti S., Hierarchy of visual features for object recognition, Proceedings of the IEEE International Conference on Image Processing, pp. 5901-5905, (2014)
123