Michael Kah Ong Goh

Wavelet local binary patterns fusion as illuminated facial image preprocessing for face verification

Poor illumination condition is recognized as one of the major problem in contemporary two-dimensional (2D) face verification system. It causes large variation in facial images and degrades the performance of the system. Many works of resolving illumination variation in face verification have been reported in the past decades. In this paper, a facial image illumination invariant technique is devised based on the fusion of wavelet analysis and local binary patterns. Particularly, illumination-reflectance model is used to detach illumination and reflectance components with multi-resolution nature of wavelet analysis. The illumination component that resides in low spatial-frequency wavelet subband is first rid off efficiently. The reflectance components that reside in high and middle spatial-frequency wavelet subbands are enhanced with local binary patterns histogram. Finally, two processed images are fused through wavelet image fusion. This technique works out promisingly in achieving better recognition results on YaleB, CMU PIE and FRGC face databases in comparison with existing illumination invariant techniques.

A Grassmann graph embedding framework for gait analysis

Gait recognition is important in a wide range of monitoring and surveillance applications. Gait information has often been used as evidence when other biometrics is indiscernible in the surveillance footage. Building on recent advances of the subspace-based approaches, we consider the problem of gait recognition on the Grassmann manifold. We show that by embedding the manifold into reproducing kernel Hilbert space and applying the mechanics of graph embedding on such manifold, significant performance improvement can be obtained. In this work, the gait recognition problem is studied in a unified way applicable for both supervised and unsupervised configurations. Sparse representation is further incorporated in the learning mechanism to adaptively harness the local structure of the data. Experiments demonstrate that the proposed method can tolerate variations in appearance for gait identification effectively.

A Grassmannian Approach to Address View Change Problem in Gait Recognition

Gait recognition appears to be a valuable asset when conventional biometrics cannot be employed. Nonetheless, recognizing human by gait is not a trivial task due to the complex human kinematic structure and other external factors affecting human locomotion. A major challenge in gait recognition is view variation. A large difference between the views in the query and reference sets often leads to performance deterioration. In this paper, we show how to generate virtual views to compensate the view difference in the query and reference sets, making it possible to match the query and reference sets using standardized views. The proposed method, which combines multiview matrix representation and a novel randomized kernel extreme learning machine, is an end-to-end solution for view change problem under Grassmann manifold treatment. Under the right condition, the view-tagging problem can be eliminated. Since the recording angle and walking direction of the subject are not always available, this is particularly valuable for a practical gait recognition system. We present several working scenarios for multiview recognition that have not be considered before. Rigorous experiments have been conducted on two challenging benchmark databases containing multiview gait datasets. Experiments show that the proposed approach outperforms several state-of-the-arts methods.

Gait recognition using Sparse Grassmannian Locality Preserving Discriminant Analysis

One of the greatest challenges for gait recognition is identification across appearance change. In this paper, we present a gait recognition method called Sparse Grassmannian Locality Preserving Discriminant Analysis. The proposed method learns a compact and rich representation of the gait images through sparse representation. The use of Grassmannian locality preserving discriminant analysis further optimizes the performance by preserving both global discriminant and local geometrical structure of the gait data. Experiments demonstrate that the proposed method can tolerate variation in appearance for gait identification effectively.

Human gait recognition using localized Grassmann mean representatives with partial least squares regression

Gait recognition has become popular due to the rising demand for nonintrusive biometrics. At its nascent stage of development, gait recognition faces a number of challenges. The performance of a gait recognition system is sensitive towards factors like viewing angle, clothing, shoe type, load carriage and speed changes. In this paper, the problems of gait are formulated on the Grassmann manifold. It is not difficult to obtain multiple snapshots of a walking subjects with the wide availability of camera networks. These sets of images can be modelled as low-dimensional subspaces, which can be realized naturally as points on the Grassmann manifold. Modelling image sets as low-dimensional subspaces provides not only possible clue of one’s gait, but also the common patterns of variation in the set. We present a method called Localized Grassmann Mean Representatives with Partial Least Squares Regression (LoGPLS) to infer a low-dimensional Euclidean approximation of the manifold. The notion of local mean representatives is introduced to construct multiple tangent spaces to better approximate the topological structure of the manifold. As the properties of the tangent spaces allows the Grassmann points to be evaluated in the vector space, partial least squares is applied to allow a more accurate classification of the points in a reduced space. Experiments have been conducted on four different publicly available gait databases. Empirical evidences demonstrate the effectiveness of the proposed approach in solving the various covariates in gait recognition.

Finger Vein verification using local histogram of hybrid texture descriptors

Lately, finger vein has been recognized as an efficacious biometric method for user authentication due to the uniqueness of vein patterns and its insusceptibility to forgery because the vein patterns reside inside the human body. In this work, hybrid histogram descriptor is the proposed method. This method utilizes the sign and magnitude components of the texture extracted by using Binary Gradient Contour (BGC). Subsequently, the histogram is locally computed to determine the weight distribution of the sign and magnitude value for the hybrid texture descriptors. The extensive experimental results demonstrate the overall superiority of the proposed method with the EER as low as 0.353%.

Multi-instance finger vein recognition using local hybrid binary gradient contour

In a finger vein authentication system, the image of a finger acquired for recognition always suffers from noises due to imperfect acquisition device, signal distortion, and variability of individual physical appearance over time. To improve the system performance, we propose a multi-instances finger vein recognition using feature level fusion. Local Hybrid Binary Gradient Contour (LHBGC) is proposed as the finger texture descriptor and SVM is used for classification. Experiments are conducted using the Shandong finger vein database (SDUMLA-HMT) and also the University Sains Malaysia finger vein database (FV-USM). Experimental results show a significant increase in performance accuracy when more than one fingers are combined, with an EER as low as 0.0038%.

Robust hybrid descriptors for multi-instance finger vein recognition

Finger vein recognition is a type of biometric technology that uses the vein pattern inside the human finger as a personal identifier. In this paper, Local Hybrid Binary Gradient Contour (LHBGC) and Hierarchical Local Binary Pattern (HLBP) are proposed as the texture descriptors for finger vein recognition to increase the discriminant capability of the finger vein texture. LHBGC extracts both sign and magnitude components of the finger vein image for recognition, while HLBP utilizes the LBP uniform texture pattern of the vein image without any training required. Furthermore, a multi-instance biometrics that fuses multiple evidences from an individual has also been proposed to address the problem of noisy data. Multi-instance biometrics is the most inexpensive way to obtain multiple biometric evidences from a biometric trait without multiple sensors and additional feature extraction algorithms. Experiments on several benchmark databases validate the efficiency of the proposed multi-instance approach. An equal error rate as low as 0.00002% is achieved using the combination of three fingers at score level fusion.

Personal Recognition Using Multi-angles Gait Sequences

This paper presents an automatic gait recognition system which recognizes a person by the way he/she walks. The gait signature is obtained based on the contour width information of the silhouette of a person. Using this statistical shape information, we could capture the compact structural and dynamic features of the walking pattern. As the extracted contour width feature is large in size, Fisher Discriminant Analysis is used to reduce the dimension of the feature set. After that, a modified Probabilistic Neural Networks is deployed to classify the reduced feature set. Satisfactory result could be achieved when we fused gait images from multiple viewing angles. In this research, we aim to identify the complete gait cycle of each subjects. Every person walks at difference paces and thus different numbers of frame sizes are required to record the walking pattern. As such, it is not robust and feasible if we take a fixed number of video frames to process the gait sequences for all subjects. We endeavor to find an efficient method to identify the complete gait cycle of each individual. In this case, we could work on succinct representation of the gait pattern which is invariant to walking speed for each individual.

Illuminated facial preprocessing with Wavelet Local Binary Patterns Histogram Specification

Contemporary 2D face recognition is still a challenging work, especially when lighting varies. Thus, many works of resolving illumination variation in face recognition have been proposed, in the past decades. In this paper, we proposed Wavelet Local Binary Patterns Histogram Specification as a preprocessing technique for illuminated face recognition. Based on wavelet analysis, an illuminated facial image is decomposed into illumination and reflectance components. The illumination component that resides in the low spatial-frequency subband is first removed. Next, the reflectance component that resides in the high and middle spatial-frequency subband is then enhanced with local binary pattern histogram. This technique is promising in achieving better recognition performance on YaleB and CMU PIE face databases in comparison to the results that achieved by existing illumination invariant techniques.