Feifei Lee

Face Recognition Using Self-Organizing Maps

As an active research area, face recognition has been studied for more than 20 years. Especially, after the September 11 terrorist attacks on the United States, security systems utilizing personal biometric features, such as, face, voice, fingerprint, iris pattern, etc. are attracting a lot of attention. Among them, face recognition systems have become the subject of increased interest (Bowyer, 2004). Face recognition seems to be the most natural and effective method to identify a person since it is the same as the way human does and there is no need to use special equipments. In face recognition, personal facial feature extraction is the key to creating more robust systems. A lot of algorithms have been proposed for solving face recognition problem. Based on the use of the Karhunen-Loeve transform, PCA (Turk & Pentland, 1991) is used to represent a face in terms of an optimal coordinate system which contains the most significant eigenfaces and the mean square error is minimal. However, it is highly complicated and computational-power hungry, making it difficult to implement them into real-time face recognition applications. Feature-based approach (Brunelli & Poggio, 1993; Wiskott et al., 1997) uses the relationship between facial features, such as the locations of eye, mouth and nose. It can implement very fast, but recognition rate usually depends on the location accuracy of facial features, so it can not give a satisfied recognition result. There are many other algorithms have been used for face recognition. Such as Local Feature Analysis (LFA) (Penev & Atick, 1996), neural network (Chellappa et al., 1995), local autocorrelations and multi-scale integration technique (Li & Jain, 2005), and other techniques (Goudail et al., 1996; Moghaddam & Pentland, 1997; Lam & Yan, 1998; Zhao, 2000; Bartlett et al., 2002 ; Kotani et al., 2002; Karungaru et al., 2005; Aly et al., 2008) have been proposed. As a neural unsupervised learning algorithm, Kohonen’s Self-Organizing Maps (SOM) has been widely utilized in pattern recognition area. In this chapter, we will give an overview in SOM-based face recognition applications. Using the SOM as a feature extraction method in face recognition applications is a promising approach, because the learning is unsupervised, no pre-classified image data are needed at all. When high compressed representations of face images or their parts are formed by the SOM, the final classification procedure can be fairly simple, needing only a moderate number of labeled training samples. In this chapter, we will introduce various face recognition algorithms based on this consideration. 17

VQ-based face recognition algorithm using code pattern classification and Self-Organizing Maps

In this paper, an improved codebook design method is proposed for VQ-based fast face recognition algorithm to improve recognition accuracy. Combined by a systematically organized codebook based on the classification of code patterns abstracted from facial images and another codebook created by Kohonenpsilas Self-Organizing Maps (SOM) method, an optimized codebook consisted of 2times2 codevectors for facial images is generated. The performance of proposed algorithm is demonstrated by using publicly available AT&T database containing variations in lighting, posing, and expressions. Compared with the algorithms employing original codebook or SOM codebook separately, experimental results show face recognition using proposed codebook is more efficient. The highest average recognition rate of 98.6% is obtained for 40 personspsila 400 images of AT&T database.

Face Recognition Using Markov Stationary Features and Vector Quantization Histogram

We have proposed a very simple yet highly reliable face recognition algorithm using VQ histogram. This histogram, obtained by Vector Quantization (VQ) processing for the facial image, is utilized as a very effective personal feature. In this paper, we combine the VQ histogram with Markov Stationary Features (MSF) so as to add spatial structure information to histogram. Experimental results show maximum average recognition rate of 96.16% is obtained for 400 images of 40 persons from the publicly available face database of AT&T Laboratories Cambridge.

A VQ-based fast face recognition algorithm using optimized codebook

In this paper, we present a VQ-based fast face recognition algorithm using an optimized codebook. Previously, Chen et al. [2006] proposed a novel codebook design method based on the systematic classification and organization of code patterns abstracted from facial images for reliable face recognition. In this paper, an improved codebook design method is proposed. Combined by a systematically organized codebook based on the classification of code patterns and another codebook created by Kohonenpsilas Self-Organizing Maps (SOM), an optimized codebook consisted of 2times2 codevectors for facial images is generated. We demonstrate the performance of our algorithm using publicly available AT&T database containing variations in lighting, posing, and expressions. Compared with the algorithms employing original codebook or SOM codebook separately, experimental results show face recognition using the optimized codebook is more efficient. The highest average recognition rate of 98.2% is obtained for 40 personspsila 400 images of AT&T database. A table look-up (TLU) method is also proposed for the speed up of the recognition processing in this paper. By applying this method in the quantization step, the total recognition processing time achieves only 28 msec, enabling real-time face recognition.

Face recognition based on the combination of histogram features and rough location information of facial parts

In this paper, an improved face recognition algorithm based on adjacent pixel intensity difference quantization (APIDQ) histogram method is proposed. By utilizing the rough location information of the facial parts, the facial area is divided into 5 individual parts, and then APIDQ is applied on each facial component. Recognition results are firstly obtained from different parts separately and then combined by weighted averaging. The experimental result shows that top 1 recognition rate of 97.6% is achieved when evaluated by FB task of the FERET database.

Region-Division Vector Quantization Histogram Method for Human Face Recognition

Abstract We have developed a very simple yet highly reliable face recognition method called VQ histogram method codevector referred (or matched) count histogram, which is obtained by Vector Quantization (VQ) processing of facial image, is utilized as a very effective personal feature value. Furthermore, for adding the geometric information of the face to improve the recognition accuracy, aregion-division (RD) VQ histogram method is proposed in this paper. We divide the facial area into 5 regions relating to the facial parts (forehead, eye, nose, mouth, jaw). Recognition results with different parts are fast obtained separately and then combined by weighted averaging. Topl recognition rate of 97.4% is obtained by using FB task (1195 images) in the standard FERET database. By using the private database, which was taken in practical but yet reasonably regulated environrnent, Topl recognition rate of 100% is realized.

Feature facial image recognition using VQ histogram in the DCT domain

In this paper, a novel algorithm using vector quantization (VQ) method for facial image recognition in DCT domain is presented. Firstly, feature vectors of facial image are generated by using DCT (Discrete Cosine transform) coefficients in low frequency domains. Then codevector referred count histogram, which is utilized as a very effective personal feature value, is obtained by Vector Quantization (VQ) processing. Publicly available AT&T database of 40 subjects with 10 images per subject containing variations in lighting, posing, and expressions, is used to evaluate the performance of the proposed algorithm. Experimental results show face recognition using proposed feature vector is very efficient. The highest average recognition rate of 94.8% is obtained.

A Codebook Design Method for Robust VQ-Based Face Recognition Algorithm

In this paper, we present a theoretical codebook design method for VQ-based fast face recognition algorithm to im-prove recognition accuracy. Based on the systematic analysis and classification of code patterns, firstly we theoretically create a systematically organized codebook. Combined with another codebook created by Kohonen’s Self-Organizing Maps (SOM) method, an optimized codebook consisted of 2×2 codevectors for facial images is generated. Experimental results show face recognition using such a codebook is more efficient than the codebook consisted of 4×4 codevector used in conventional algorithm. The highest average recognition rate of 98.6% is obtained for 40 persons’ 400 images of publicly available face database of AT&T Laboratories Cambridge containing variations in lighting, posing, and expressions. A table look-up (TLU) method is also proposed for the speed up of the recognition processing. By applying this method in the quantization step, the total recognition processing time achieves only 28 msec, enabling real-time face recognition.

An improved fast face recognition algorithm based on adjacent pixel intensity difference quantization histogram

In this paper, we present an improved face recognition algorithm based on adjacent pixel intensity difference quantization (APIDQ) histogram method proposed by Kotani et al. [12]. We optimize the quantization method of APIDQ according to the maximum entropy principle (MEP), and determine the best parameters for APIDQ. Experimental results show maximum average recognition rate of 97.2% for 400 images of 40 persons (10 images per person) from the publicly available AT&T face database.

Fast Search Algorithm for Short Video Clips from Large Video Database Using a Novel Histogram Feature

In this paper, we present a novel fast video search algorithm for large video database. This algorithm is based on the adjacent pixel intensity difference quantization (APIDQ) algorithm, which had been reliably applied to human face recognition previously. An APIDQ histogram is utilized as the feature vector of the frame image. Combined with active search, a temporal pruning algorithm, fast and robust video search can be achieved. The proposed search algorithm has been evaluated by 6 hours of video to search for given 200 video clips which each length is 15 seconds. Experimental results show the proposed algorithm can detect the similar video clip in merely 80 ms, and is more accurately and robust against Gaussian noise than conventional fast video search algorithm.