期刊文献+
任意字段
题名或关键词
题名
关键词
文摘
作者
第一作者
机构
刊名
分类号
参考文献
作者简介
基金资助
栏目信息

改进卷积玻尔兹曼机的图像特征深度提取 被引量：11

New image deep feature extraction based on improved CRBM
下载PDF
导出
摘要 针对卷积深度和信念网络存在计算复杂度高和训练缓慢的问题,提出卷积深度玻尔兹曼机用于图像特征提取.针对卷积受限玻尔兹曼机进行改进,提出最大化图像中间区域概率的训练目标函数,并引入性能较好的交叉熵稀疏惩罚因子和dropout训练方法.设计卷积深度玻尔兹曼机结构,提出均值聚合机制,将聚合层内点的值定义为block中各点激活概率均值,对层间关联进行简化,将聚合层内各面直接叠加以供高层CRBM提取特征.通过在MNIST手写数字识别集上的实验结果证明,采用新模型提取的图像特征分类准确率提高0.5%、训练时间减少50%,且达到了目前MNIST数据集的最佳水平. To resolve the problems of high computational complexity and slow training in Convolutional Deep Belief Net, Convolutional Deep Bohzmann Machine （CDBM） is proposed to extract image features. To improve the Convelution Restricted Boltzmann Machine（ CRBM）, a new training objective function to maximize the probability of intermediate image area is proposed, along with introducing the cross-entropy penalty factor and dropout training. After that, CDBM is designed based on modified CRBM. The mean-pool mechanism is presented to lessen computational complexity and improve the robustness of features for image scaling. The relationship between layers is simplified to extract high-level abstract features. The MNIST handwritten digits database is used to test this new model and the results prove that features extracted by CDBM are more accurate than CDBN. The classification accuracy rate increase at least 0.5%, and training time decrease more than 50%.
作者 刘凯 张立民 范晓磊
机构地区 海军航空工程学院基础实验部 海军航空工程学院信息融合研究所 第二炮兵工程大学士官职业技术教育学院
出处 《哈尔滨工业大学学报》 EI CAS CSCD 北大核心 2016年第5期155-159,共5页 Journal of Harbin Institute of Technology
基金 国家自然科学基金项目(61032001)
关键词 深度学习 图像特征提取 卷积受限玻尔兹曼机 卷积深度玻尔兹曼机 deep learning image features extraction CRBM CDBM
分类号 TP391.4 [自动化与计算机技术—计算机应用技术]
  • 相关文献

参考文献14

  • 1刘建伟,刘媛,罗雄麟.玻尔兹曼机研究进展[J].计算机研究与发展,2014,51(1):1-16. 被引量：71
  • 2LAROCHELLE H, ERHAN D, COURVILLE A, et al. An empirical evaluation of deep architectures on problems with many factors of variation [ C ]// Proceedings of the 24th international conference on Machine learning. New York: ACM, 2007: 473-480.
  • 3SRIVASTAVA N, SALAKHUTDINOV R. Multimodal learning with deep boltzmann machines[ C]// Advances in neural information processing systems 2012. New York: ACM, 2012: 2222-2230.
  • 4LEE H, GROSSE R, RANGANATH Convolutional deep belief networks R, et al. for scalable unsupervised learning of hierarchical representations [ C ]//Proceedings of the 26th Annual International Conference on Machine Learning. New York: ACM, 2009: 609-616.
  • 5GOODFELLOW I, WARDE-FARLEY D, MIRZA M, et al. Maxout Networks [ C ]// Proceedings of the 30th International Conference on Machine Learning. Piscataway. NJ: IEEE, 2013: 1319-1327.
  • 6CIRESAN D C, MEIER U, MASCI J, et al. Flexible, high performance convolutional neural networks for image classification [ C ]// Proceedings International Joint Conference on Artificial Intelligence 2011. New York: ACM, 2011: 22(1) . 1237-1239.
  • 7CIRESAN D, MEIER U, SCHMIDHUBER J. Multi-column deep neural networks for image classification [ C ]// IEEE Conference on Computer Vision and Pattern Recognition 2012. Piscataway, NJ: IEEE, 2012: 3642-3649.
  • 8LEE H, EKANADHAM C, NG A Y. Sparse deep belief net model for visual area V2 [ C ]// Advances in Neural Information Processing Systems 2008 . New York: ACM, 2008 : 873-880.
  • 9HINTON G. A practical guide to training restricted boltzmann machines [ M ]. Berlin : Springer Berlin Heidelberg, 2012: 599-619.
  • 10LAWRENCE S, GILES C L, TSOI A C, et al. Face recognition: A convolutional neural-network approach [ J ]. IEEE Transactions on Neural Networks, 1997, 8( 1 ) : 98- 113.

二级参考文献46

  • 1Hopfield J J. Neural networks and physical systems with emergent col|ective computational abilities [J]. Proceedings of the National Academy of Sciences of the United States of America, 1982, 79(8):2554-2558.
  • 2Hinton G E, Sejnowski T J. Optimal perceptual inference [C]//Proc of the 1983 IEEE Conf on Computer Vision and Pattern Recognition. Los Alamitos, CA: IEEE Computer Society, 1983: 448-453.
  • 3Hinton G E, Sejnowski T J. Analyzing cooperative computation [C] //Proc of the 5th Annual Congress of the Cognitive Science Society. New York: ACM, 1983: 2554- 2558.
  • 4Hofstadter D R. The copycat project: An experiment in nondeterminism and creative analogies [DB/OL:. MIT Artificial Intelligence Laboratory Memo 755. (1984- 01-01) [2004-10-01]. http://hdl, handle, net/1721.1/5648.
  • 5Hofstadter D R. A Non-Deterministic Approach to Analogy, Involving the Ising Model of Ferromagnetism[M] //The Physics of Cognitive Processes. Hackensack: World Scientific, 1987.
  • 6Smolensky P. Information Processing in Dynamical Systems: Foundations of Harmony Theory [M]//Parallel Distributed Processing, Vol 1: Foundations. Cambridge: MIT Press, 1986: 194-281.
  • 7Ackley D H, Hinton G E, Sejnowski T J. A learning algorithm for Boltzmann machines [J]. Cognitive Science, 1985, 9(1): 147-169.
  • 8Hinton G E. Training products otF experts by minimizing contrastive divergence[J]. Neural Computation, 2002, 14 (8) : 1771-1800.
  • 9Kirkpatrick S, Gelatt C D, Vecchi M P. Optimization by simulated annealing [J]. Science, 1983, 220 (4598) : 671- 680.
  • 10Hinton G E. To recognize shapes, first learn to generate images [J]. Computational Neuroscience: Theoretical Insights into Brain Function, 2007, 165(1): 535-547.

共引文献70

同被引文献55

引证文献11

二级引证文献43

哈尔滨工业大学学报
2016年 第5期

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部