Search For:

Displaying 1-16 out of 16 total
Supervised Self-taught Learning: Actively transferring knowledge from unlabeled data
Found in: Neural Networks, IEEE - INNS - ENNS International Joint Conference on
By Kaizhu Huang, Zenglin Xu, Irwin King, Michael R. Lyu, Colin Campbell
Issue Date:June 2009
pp. 1272-1277
We consider the task of Self-taught Learning (STL) from unlabeled data. In contrast to semi-supervised learning, which requires unlabeled data to have the same set of class labels as labeled data, STL can transfer knowledge from different types of unlabele...
 
Robust Text Detection in Natural Scene Images
Found in: IEEE Transactions on Pattern Analysis and Machine Intelligence
By Xu-Cheng Yin,Xuwang Yin,Kaizhu Huang,Hong-Wei Hao
Issue Date:May 2014
pp. 1-1
Text detection in natural scene images is an important prerequisite for many content-based image analysis tasks. In this paper, we propose an accurate and robust method for detecting texts in natural scene images. A fast and effective pruning algorithm is ...
 
Feature Transformation with Class Conditional Decorrelation
Found in: 2013 IEEE International Conference on Data Mining (ICDM)
By Xu-Yao Zhang,Kaizhu Huang,Cheng-Lin Liu
Issue Date:December 2013
pp. 887-896
The well-known feature transformation model of Fisher linear discriminant analysis (FDA) can be decomposed into an equivalent two-step approach: whitening followed by principal component analysis (PCA) in the whitened space. By proving that whitening is th...
 
Low Rank Metric Learning with Manifold Regularization
Found in: Data Mining, IEEE International Conference on
By Guoqiang Zhong,Kaizhu Huang,Cheng-Lin Liu
Issue Date:December 2011
pp. 1266-1271
In this paper, we present a semi-supervised method to learn a low rank Mahalanobis distance function. Based on an approximation to the projection distance from a manifold, we propose a novel parametric manifold regularizer. In contrast to previous approach...
 
Fast and Robust Graph-based Transductive Learning via Minimum Tree Cut
Found in: Data Mining, IEEE International Conference on
By Yan-Ming Zhang,Kaizhu Huang,Cheng-Lin Liu
Issue Date:December 2011
pp. 952-961
In this paper, we propose an efficient and robust algorithm for graph-based transductive classification. After approximating a graph with a spanning tree, we develop a linear-time algorithm to label the tree such that the cut size of the tree is minimized....
 
Similar Handwritten Chinese Characters Recognition by Critical Region Selection Based on Average Symmetric Uncertainty
Found in: Frontiers in Handwriting Recognition, International Conference on
By Bo Xu, Kaizhu Huang, Cheng-Lin Liu
Issue Date:November 2010
pp. 527-532
We consider the problem of similar Chinese character recognition in this paper. Engaging the Average Symmetric Uncertainty (ASU) criterion to measure the correlation between different image regions and the class label, we manage to detect the most critical...
 
Dimensionality Reduction by Minimal Distance Maximization
Found in: Pattern Recognition, International Conference on
By Bo Xu, Kaizhu Huang, Cheng-Lin Liu
Issue Date:August 2010
pp. 569-572
In this paper, we propose a novel discriminant analysis method, called Minimal Distance Maximization (MDM). In contrast to the traditional LDA, which actually maximizes the average divergence among classes, MDM attempts to find a low-dimensional subspace t...
 
GSML: A Unified Framework for Sparse Metric Learning
Found in: Data Mining, IEEE International Conference on
By Kaizhu Huang, Yiming Ying, Colin Campbell
Issue Date:December 2009
pp. 189-198
There has been significant recent interest in sparse metric learning (SML) in which we simultaneously learn both a good distance metric and a low-dimensional representation. Unfortunately, the performance of existing sparse metric learning approaches is us...
 
Semi-supervised Learning from General Unlabeled Data
Found in: Data Mining, IEEE International Conference on
By Kaizhu Huang, Zenglin Xu, Irwin King, Michael R. Lyu
Issue Date:December 2008
pp. 273-282
We consider the problem of Semi-supervised Learning (SSL) from general unlabeled data, which may contain irrelevant samples. Within the binary setting, our model manages to better utilize the information from unlabeled data by formulating them as a three-c...
 
Direct Zero-Norm Optimization for Feature Selection
Found in: Data Mining, IEEE International Conference on
By Kaizhu Huang, Irwin King, Michael R. Lyu
Issue Date:December 2008
pp. 845-850
Zero-norm, defined as the number of non-zero elements in a vector, is an ideal quantity for feature selection. However, minimization of zero-norm is generally regarded as a combinatorially difficult optimization problem. In contrast to previous methods tha...
 
A Scenario-View Based Approach to Analyze External Behavior of Web Services for Supporting Mediated Service Interactions
Found in: Services Computing, IEEE International Conference on
By Zhangbing Zhou, Sami Bhiri, Lei Shu, Kaizhu Huang, Laurentiu Vasiliu, Manfred Hauswirth
Issue Date:July 2008
pp. 249-256
Web service interactions have triggered the initiative to identify and solve mismatches from a behavioral aspect. Current approaches are limited since they mainly focus on control-flow but largely ignore data-flow. In this paper, we propose an approach to ...
 
An SVM-Based High-accurate Recognition Approach for Handwritten Numerals by Using Difference Features
Found in: Document Analysis and Recognition, International Conference on
By Kaizhu Huang, Jun Sun, Yoshinobu Hotta, Katsuhito Fujimoto, Satoshi Naoi
Issue Date:September 2007
pp. 589-593
Handwritten numeral recognition is an important pattern recognition task. It can be widely used in various domains, e.g., bank money recognition, which requires a very high recognition rate. As a state-of-the-art classifier, Support Vector Machine (SVM), h...
 
Learning Classifiers from Imbalanced Data Based on Biased Minimax Probability Machine
Found in: Computer Vision and Pattern Recognition, IEEE Computer Society Conference on
By Kaizhu Huang, Haiqin Yang, Irwin King, Michael R. Lyu
Issue Date:July 2004
pp. 558-563
We consider the problem of the binary classification on imbalanced data, in which nearly all the instances are labelled as one class, while far fewer instances are labelled as the other class, usually the more important class. Traditional machine learning ...
 
Accurate and robust text detection: a step-in for text retrieval in natural scene images
Found in: Proceedings of the 36th international ACM SIGIR conference on Research and development in information retrieval (SIGIR '13)
By Hong-Wei Hao, Kaizhu Huang, Xu-Cheng Yin, Xuwang Yin
Issue Date:July 2013
pp. 1091-1092
We propose and implement a robust text detection system, which is a prominent step-in for text retrieval in natural scene images or videos. Our system includes several key components: (1) A fast and effective pruning algorithm is designed to extract Maxima...
     
Semi-supervised text categorization by active search
Found in: Proceeding of the 17th ACM conference on Information and knowledge mining (CIKM '08)
By Irwin King, Kaizhu Huang, Michael R. Lyu, Rong Jin, Zenglin Xu
Issue Date:October 2008
pp. 1001-1001
In automated text categorization, given a small number of labeled documents, it is very challenging, if not impossible, to build a reliable classifier that is able to achieve high classification accuracy. To address this problem, a novel web-assisted text ...
     
Learning large margin classifiers locally and globally
Found in: Twenty-first international conference on Machine learning (ICML '04)
By Haiqin Yang, Irwin King, Kaizhu Huang, Michael R. Lyu
Issue Date:July 2004
pp. 182-182
A new large margin classifier, named Maxi-Min Margin Machine (M4) is proposed in this paper. This new classifier is constructed based on both a "local: and a "global" view of data, while the most popular large margin classifier, Support Vector Machine (SVM...
     
 1