Time filter

Source Type

Ma A.J.,Hong Kong Baptist University | Yuen P.C.,Hong Kong Baptist University | Yuen P.C.,PBNU HKBU United International College | Zou W.W.W.,Hong Kong Baptist University | And 2 more authors.
IEEE Transactions on Circuits and Systems for Video Technology

Supervised manifold learning has been successfully applied to action recognition, in which class label information could improve the recognition performance. However, the learned manifold may not be able to well preserve both the local structure and global constraint of temporal labels in action sequences. To overcome this problem, this paper proposes a new supervised manifold learning algorithm called supervised spatio-temporal neighborhood topology learning (SSTNTL) for action recognition. By analyzing the topological characteristics in the context of action recognition, we propose to construct the neighborhood topology using both supervised spatial and temporal pose correspondence information. Employing the property in locality preserving projection (LPP), SSTNTL solves the generalized eigenvalue problem to obtain the best projections that not only separates data points from different classes, but also preserves local structures and temporal pose correspondence of sequences from the same class. Experimental results demonstrate that SSTNTL outperforms the manifold embedding methods with other topologies or local discriminant information. Moreover, compared with state-of-the-art action recognition algorithms, SSTNTL gives convincing performance for both human and gesture action recognition. © 1991-2012 IEEE. Source

Ma A.J.,Hong Kong Baptist University | Yuen P.C.,Hong Kong Baptist University | Yuen P.C.,BNU HKBU United International College | Lai J.-H.,Sun Yat Sen University | Lai J.-H.,Guangdong Province Key Laboratory of Information Security
IEEE Transactions on Pattern Analysis and Machine Intelligence

This paper addresses the independent assumption issue in fusion process. In the last decade, dependency modeling techniques were developed under a specific distribution of classifiers or by estimating the joint distribution of the posteriors. This paper proposes a new framework to model the dependency between features without any assumption on feature/classifier distribution, and overcomes the difficulty in estimating the high-dimensional joint density. In this paper, we prove that feature dependency can be modeled by a linear combination of the posterior probabilities under some mild assumptions. Based on the linear combination property, two methods, namely, Linear Classifier Dependency Modeling (LCDM) and Linear Feature Dependency Modeling (LFDM), are derived and developed for dependency modeling in classifier level and feature level, respectively. The optimal models for LCDM and LFDM are learned by maximizing the margin between the genuine and imposter posterior probabilities. Both synthetic data and real datasets are used for experiments. Experimental results show that LCDM and LFDM with dependency modeling outperform existing classifier level and feature level combination methods under nonnormal distributions and on four real databases, respectively. Comparing the classifier level and feature level fusion methods, LFDM gives the best performance. © 1979-2012 IEEE. Source

Wu J.-S.,Sun Yat Sen University | Wu J.-S.,SYSU CMU Shunde International Joint Research Institute | Zheng W.-S.,Sun Yat Sen University | Zheng W.-S.,Guangdong Province Key Laboratory of Computational Science | And 2 more authors.
Neural Networks

Kernel competitive learning has been successfully used to achieve robust clustering. However, kernel competitive learning (KCL) is not scalable for large scale data processing, because (1) it has to calculate and store the full kernel matrix that is too large to be calculated and kept in the memory and (2) it cannot be computed in parallel. In this paper we develop a framework of approximate kernel competitive learning for processing large scale dataset. The proposed framework consists of two parts. First, it derives an approximate kernel competitive learning (AKCL), which learns kernel competitive learning in a subspace via sampling. We provide solid theoretical analysis on why the proposed approximation modelling would work for kernel competitive learning, and furthermore, we show that the computational complexity of AKCL is largely reduced. Second, we propose a pseudo-parallelled approximate kernel competitive learning (PAKCL) based on a set-based kernel competitive learning strategy, which overcomes the obstacle of using parallel programming in kernel competitive learning and significantly accelerates the approximate kernel competitive learning for large scale clustering. The empirical evaluation on publicly available datasets shows that the proposed AKCL and PAKCL can perform comparably as KCL, with a large reduction on computational cost. Also, the proposed methods achieve more effective clustering performance in terms of clustering precision against related approximate clustering approaches. © 2014 Elsevier Ltd. Source

Xie X.,Sun Yat Sen University | Zheng W.-S.,Sun Yat Sen University | Zheng W.-S.,Queen Mary, University of London | Lai J.,Sun Yat Sen University | And 3 more authors.
IEEE Transactions on Image Processing

A face image can be represented by a combination of large-and small-scale features. It is well-known that the variations of illumination mainly affect the large-scale features (low-frequency components), and not so much the small-scale features. Therefore, in relevant existing methods only the small-scale features are extracted as illumination-invariant features for face recognition, while the large-scale intrinsic features are always ignored. In this paper, we argue that both large-and small-scale features of a face image are important for face restoration and recognition. Moreover, we suggest that illumination normalization should be performed mainly on the large-scale features of a face image rather than on the original face image. A novel method of normalizing both the Small-and Large-scale (S&L) features of a face image is proposed. In this method, a single face image is first decomposed into large-and small-scale features. After that, illumination normalization is mainly performed on the large-scale features, and only a minor correction is made on the small-scale features. Finally, a normalized face image is generated by combining the processed large-and small-scale features. In addition, an optional visual compensation step is suggested for improving the visual quality of the normalized image. Experiments on CMU-PIE, Extended Yale B, and FRGC 2.0 face databases show that by using the proposed method significantly better recognition performance and visual results can be obtained as compared to related state-of-the-art methods. © 2011 IEEE. Source

Xie X.,Sun Yat Sen University | Xie X.,Guangdong Province Key Laboratory of Information Security | Lai J.,Sun Yat Sen University | Lai J.,Guangdong Province Key Laboratory of Information Security | Zheng W.-S.,Queen Mary, University of London
Pattern Recognition

Face recognition under varying lighting conditions is challenging, especially for single image based recognition system. Exacting illumination invariant features is an effective approach to solve this problem. However, existing methods are hard to extract both multi-scale and multi-directivity geometrical structures at the same time, which is important for capturing the intrinsic features of a face image. In this paper, we propose to utilize the logarithmic nonsubsampled contourlet transform (LNSCT) to estimate the reflectance component from a single face image and refer it as the illumination invariant feature for face recognition, where NSCT is a fully shift-invariant, multi-scale, and multi-direction transform. LNSCT can extract strong edges, weak edges, and noise from a face image using NSCT in the logarithm domain. We analyze that in the logarithm domain the low-pass subband of a face image and the low frequency part of strong edges can be regarded as the illumination effects, while the weak edges and the high frequency part of strong edges can be considered as the reflectance component. Moreover, even though a face image is polluted by noise (in particular the multiplicative noise), the reflectance component can still be well estimated and meanwhile the noise is removed. The LNSCT can be applied flexibly as neither assumption on lighting condition nor information about 3D shape is required. Experimental results show the promising performance of LNSCT for face recognition on Extended Yale B and CMU-PIE databases. © 2010 Elsevier Ltd. ALL rights reserved. Source

Discover hidden collaborations