Title :
Discriminative Deep Belief Networks for image classification
Author :
Zhou, Shusen ; Chen, Qingcai ; Wang, Xiaolong
Author_Institution :
Shenzhen Grad. Sch., Harbin Inst. of Technol., Harbin, China
Abstract :
This paper presents a novel semi-supervised learning algorithm called Discriminative Deep Belief Networks (DDBN), to address the image classification problem with limited labeled data. We first construct a new deep architecture for classification using a set of Restricted Boltzmann Machines (RBM). The parameter space of the deep architecture is initially determined using labeled data together with abundant of unlabeled data, by greedy layer-wise unsupervised learning. Then, we fine-tune the whole deep networks using an exponential loss function to maximize the separability of the labeled data, by gradient-descent based supervised learning. Experiments on the artificial dataset and real image datasets show that DDBN outperforms most semi-supervised algorithm and deep learning techniques, especially for the hard classification tasks.
Keywords :
Boltzmann machines; belief networks; image classification; learning (artificial intelligence); discriminative deep belief network; greedy layerwise unsupervised learning; image classification; restricted Boltzmann machines; semisupervised learning algorithm; Artificial neural networks; Classification algorithms; Error analysis; Learning; Supervised learning; Support vector machines; Training; Discriminative Deep Belief Networks (DDBN); deep learning; image classification; semi-supervised learning;
Conference_Titel :
Image Processing (ICIP), 2010 17th IEEE International Conference on
Conference_Location :
Hong Kong
Print_ISBN :
978-1-4244-7992-4
Electronic_ISBN :
1522-4880
DOI :
10.1109/ICIP.2010.5649922