Wenming Zheng

Locally nearest neighbor classifiers for pattern classification

In this paper, two novel classifiers based on locally nearest neighborhood rule, called nearest neighbor line and nearest neighbor plane, are presented for pattern classification. Comparison to nearest feature line and nearest feature plane, the proposed methods take much lower computation cost and achieve competitive performance.

Gender classification based on boosting local binary pattern

This paper presents a novel approach for gender classification by boosting local binary pattern-based classifiers. The face area is scanned with scalable small windows from which Local Binary Pattern (LBP) histograms are obtained to effectively express the local feature of a face image. The Chi square distance between corresponding Local Binary Pattern histograms of sample image and template is used to construct weak classifiers pool. Adaboost algorithm is applied to build the final strong classifiers by selecting and combining the most useful weak classifiers. In addition, two experiments are made for classifying gender based on local binary pattern. The male and female images set are collected from FERET databases. In the first experiment, the features are extracted by LBP histograms from fixed sub windows. The second experiment is tested on our boosting LBP based method. Finally, the results of two experiments show that the features extracted by LBP operator are discriminative for gender classification and our proposed approach achieves better performance of classification than several others methods.

L1-Norm-Based Common Spatial Patterns

Common spatial patterns (CSP) is a commonly used method of spatial filtering for multichannel electroencephalogram (EEG) signals. The formulation of the CSP criterion is based on variance using L2-norm, which implies that CSP is sensitive to outliers. In this paper, we propose a robust version of CSP, called CSP-L1, by maximizing the ratio of filtered dispersion of one class to the other class, both of which are formulated by using L1-norm rather than L2-norm. The spatial filters of CSP-L1 are obtained by introducing an iterative algorithm, which is easy to implement and is theoretically justified. CSP-L1 is robust to outliers. Experiment results on a toy example and datasets of BCI competitions demonstrate the efficacy of the proposed method.

Fisher Discriminant Analysis With L1-Norm

Fisher linear discriminant analysis (LDA) is a classical subspace learning technique of extracting discriminative features for pattern recognition problems. The formulation of the Fisher criterion is based on the L2-norm, which makes LDA prone to being affected by the presence of outliers. In this paper, we propose a new method, termed LDA-L1, by maximizing the ratio of the between-class dispersion to the within-class dispersion using the L1-norm rather than the L2-norm. LDA-L1 is robust to outliers, and is solved by an iterative algorithm proposed. The algorithm is easy to be implemented and is theoretically shown to arrive at a locally maximal point. LDA-L1 does not suffer from the problems of small sample size and rank limit as existed in the conventional LDA. Experiment results of image recognition confirm the effectiveness of the proposed method.

Locality-Preserved Maximum Information Projection

Dimensionality reduction is usually involved in the domains of artificial intelligence and machine learning. Linear projection of features is of particular interest for dimensionality reduction since it is simple to calculate and analytically analyze. In this paper, we propose an essentially linear projection technique, called locality-preserved maximum information projection (LPMIP), to identify the underlying manifold structure of a data set. LPMIP considers both the within-locality and the between-locality in the processing of manifold learning. Equivalently, the goal of LPMIP is to preserve the local structure while maximize the out-of-locality (global) information of the samples simultaneously. Different from principal component analysis (PCA) that aims to preserve the global information and locality-preserving projections (LPPs) that is in favor of preserving the local structure of the data set, LPMIP seeks a tradeoff between the global and local structures, which is adjusted by a parameter alpha, so as to find a sub- space that detects the intrinsic manifold structure for classification tasks. Computationally, by constructing the adjacency matrix, LPMIP is formulated as an eigenvalue problem. LPMIP yields orthogonal basis functions, and completely avoids the singularity problem as it exists in LPP. Further, we develop an efficient and stable LPMIP/QR algorithm for implementing LPMIP, especially, on high-dimensional data set. Theoretical analysis shows that conventional linear projection methods such as (weighted) PCA, maximum margin criterion (MMC), linear discriminant analysis (LDA), and LPP could be derived from the LPMIP framework by setting different graph models and constraints. Extensive experiments on face, digit, and facial expression recognition show the effectiveness of the proposed LPMIP method.

Foley-Sammon optimal discriminant vectors using kernel approach

A new nonlinear feature extraction method called kernel Foley-Sammon optimal discriminant vectors (KFSODVs) is presented in this paper. This new method extends the well-known Foley-Sammon optimal discriminant vectors (FSODVs) from linear domain to a nonlinear domain via the kernel trick that has been used in support vector machine (SVM) and other commonly used kernel-based learning algorithms. The proposed method also provides an effective technique to solve the so-called small sample size (SSS) problem which exists in many classification problems such as face recognition. We give the derivation of KFSODV and conduct experiments on both simulated and real data sets to confirm that the KFSODV method is superior to the previous commonly used kernel-based learning algorithms in terms of the performance of discrimination.

An efficient algorithm to solve the small sample size problem for LDA

In this paper, we present an efficient algorithm to solve the most discriminant vectors of LDA for high-dimensional data set. The experiments on ORL face database confirm the effectiveness of the proposed method.

Emotion recognition from arbitrary view facial images

Emotion recognition from facial images is a very active research topic in human computer interaction (HCI). However, most of the previous approaches only focus on the frontal or nearly frontal view facial images. In contrast to the frontal/nearly-frontal view images, emotion recognition from non-frontal view or even arbitrary view facial images is much more difficult yet of more practical utility. To handle the emotion recognition problem from arbitrary view facial images, in this paper we propose a novel method based on the regional covariance matrix (RCM) representation of facial images. We also develop a new discriminant analysis theory, aiming at reducing the dimensionality of the facial feature vectors while preserving the most discriminative information, by minimizing an estimated multiclass Bayes error derived under the Gaussian mixture model (GMM). We further propose an efficient algorithm to solve the optimal discriminant vectors of the proposed discriminant analysis method. We render thousands of multi-view 2D facial images from the BU-3DFE database and conduct extensive experiments on the generated database to demonstrate the effectiveness of the proposed method. It is worth noting that our method does not require face alignment or facial landmark points localization, making it very attractive.

Local Temporal Common Spatial Patterns for Robust Single-Trial EEG Classification

In this paper, we propose a novel optimal spatio-temporal filter, termed local temporal common spatial patterns (LTCSP), for robust single-trial elctroencephalogram (EEG) classification. Different from classical common spatial patterns (CSP) that uses only global spatial covariances to compute the optimal filter, LTCSP considers temporally local information in the variance modelling. The underlying manifold variances of EEG signals contain more discriminative information. LTCSP is an extension to CSP in the sense that CSP can be derived from LTCSP under a special case. By constructing an adjacency matrix, LTCSP is formulated as an eigenvalue problem. So, LTCSP is computationally as straightforward as CSP. However, LTCSP has better discrimination ability than CSP and is much more robust. Simulated experiment and real EEG classification demonstrate the effectiveness of the proposed LTCSP method.

Multi-View Facial Expression Recognition Based on Group Sparse Reduced-Rank Regression

In this paper, a novel multi-view facial expression recognition method is presented. Different from most of the facial expression methods that use one view of facial feature vectors in the expression recognition, we synthesize multi-view facial feature vectors and combine them to this goal. In the facial feature extraction, we use the grids with multi-scale sizes to partition each facial image into a set of sub regions and carry out the feature extraction in each sub region. To deal with the prediction of expressions, we propose a novel group sparse reduced-rank regression (GSRRR) model to describe the relationship between the multi-view facial feature vectors and the corresponding expression class label vectors. The group sparsity of GSRRR enables us to automatically select the optimal sub regions of a face that contribute most to the expression recognition. To solve the optimization problem of GSRRR, we propose an efficient algorithm using inexact augmented Lagrangian multiplier (ALM) approach. Finally, we conduct extensive experiments on both BU-3DFE and Multi-PIE facial expression databases to evaluate the recognition performance of the proposed method. The experimental results confirm better recognition performance of the proposed method compared with the state of the art methods.