Restricted boltzmann machines with SVM for object recognition

被引：0
作者：
机构：
[1] School of Mathematics and Computer Engineering, Xihua University, Chengdu
[2] Center for Radio Administration & Technology Development, Xihua University, Chengdu
来源：
Xie, Chunzhi | 1600年 / Binary Information Press卷 / 10期
基金：
中国国家自然科学基金;
关键词：
Object recognition; RBM; SVM; Unsupervised learning;
D O I：
10.12733/jcis12276
中图分类号：
学科分类号：
摘要：
Restricted Boltzmann machine (RBM) has shown the strong ability to perform unsupervised learning. Object recognition is a meaningful work under the situations where RBM approximates the stationary distribution. In this paper, RBM with SVM for object recognition is proposed. This method obtains the reduced features of data by RBM learning under the approximating distribution, and then explores SVM predicting where the reduced features coming from to estimate the classification ability of hidden states in RBM. Experiments on Yale and MNIST show that compared to some traditional methods, the proposed algorithm not only preserve the powerful learning ability of RBM but also achieve a higher recognition rate. Copyright © 2014 Binary Information Press.
引用
收藏
页码：9199 / 9206
页数：7
相关论文
共 16 条
  • [1] Hinton G.E., Salakhutdinov R.R., Reducing the dimensionality of data with neural networks, Science, 313, 5786, pp. 504-507, (2006)
  • [2] Sun Y., Wang X., Tang X., Deep Convolutional Network Cascade for Facial Point Detection, CVPR, pp. 3476-3483, (2013)
  • [3] Kohli P., Osokin A., Jegelka S., A Principled Deep Random Field Model for Image Segmentation, CVPR, pp. 1971-1978, (2013)
  • [4] Dahl G., Yu D., Deng L., Acero A., Context-dependent pre-trained deep neural networks for large vocabulary speech recognition, IEEE Trans. Speech and Audio Proc, 20, 1, pp. 30-42, (2012)
  • [5] Deng L., Hinton G.E., Kingsbury B., New types of deep neural network learning for speech recognition and related applications: an overview, ICASSP, pp. 8599-8603, (2013)
  • [6] Socher R., Perelygin A., Wu J., Chuang J., Manning C., Ng A., Potts C., Recursive Deep Models for Semantic Compositionality Over a Sentiment Treebank, EMNLP, (2013)
  • [7] Chen D., Socher R., Manning C.D., Ng A.Y., Learning New Facts From Knowledge Bases With Neural Tensor Networks and Semantic Word Vectors, ICLR, (2013)
  • [8] Salakhutdinov R., Mnih A., Hinton G.E., Restricted Boltzmann machines for collaborative filtering, ICML, pp. 791-798, (2007)
  • [9] Hinton G.E., A Practical Guide to Training Restricted Boltzmann Machines, Neural Networks: Tricks of the Trade (2nd Ed), pp. 599-619, (2012)
  • [10] Chen P.H., Lin C.J., Scholkopf B., A tutorial on -support vector machines, Appl. Stoch. Models. Bus. Ind, pp. 111-136, (2005)
12