A novel learning approach for human face detection using a network of linear units is presented. The SNoW learning architecture is a sparse network of linear functions over a pre-defined or incrementally learned feature space and is specifically tailored for learning in the presence of a very large number of features. A wide range of face images in different poses, with different expressions and under different lighting conditions are used as a training set to capture the variations of human faces. Experimental results on commonly used benchmark data sets of a wide range of face images show that the SNoW-based approach outperforms methods that use neural networks, Bayesian methods, support vector machines and others. Furthermore, learning and evaluation using the SNoW-based method are significantly more efficient than with other methods.

Two critical performance characterizations of biometric algorithms, including face recognition, are identification and verification. In face recognition, FERET is the de facto standard evaluation methodology.Identification performance of face recognition algorithms on the FERET tests has been previously reported. In this paper we report on verification performance obtained from the Sep96 FERET test. Results are presented for images taken on the same day, for images taken on different days, for images taken at least one year apart, and for images taken under different lighting conditions.

We present a technique for facial feature localization using a two-level hierarchical wavelet network. The first level wavelet network is used for face matching, and yields an affine transformation used for a rough approximation of feature locations. Second level wavelet networks for each feature are then used to fine-tune the feature locations.

We present two methods using mixtures of linear subspaces for face detection in gray level images. One method uses a mixture of factor analyzers to concurrently perform clustering and, within each cluster, perform local dimensionality reduction. The parameters of the mixture model are estimated using an EM algorithm. A face is detected if the probability of an input sample is above a predened threshold. The other mixture of subspaces method uses Kohonen 's self-organizing map for clustering and Fisher Linear Discriminant to nd an optimal projection and a Gaussian distribution to model the class-conditional density function of the projected samples for each class. The parameters of the class-conditional density functions are maximum likelihood estimates and the decision rule is also based on maximum likelihood. A wide range of face images including ones in dierent poses, with dierent expressions and under dierent lighting conditions are used as the training set to capture the varia...

Abstract. Subspace face recognition often suffers from two problems: (1) the training sample set is small compared with the high dimensional feature vector; (2) the performance is sensitive to the subspace dimension. Instead of pursuing a single optimal subspace, we develop an ensemble learning framework based on random sampling on all three key components of a classification system: the feature space, training samples, and subspace parameters. Fisherface and Null Space LDA (N-LDA) are two conventional approaches to address the small sample size problem. But in many cases, these LDA classifiers are overfitted to the training set and discard some useful discriminative information. By analyzing different overfitting problems for the two kinds of LDA classifiers, we use random subspace and bagging to improve them respectively. By random sampling on feature vectors and training samples, multiple stabilized Fisherface and N-LDA classifiers are constructed and the two groups of complementary classifiers are integrated using a fusion rule, so nearly all the discriminative information is preserved. In addition, we further apply random sampling on parameter selection in order to overcome the difficulty of selecting optimal parameters in our algorithms. Then, we use the developed random sampling framework for the integration of multiple features. A robust random sampling face recognition system integrating shape, texture, and Gabor responses is finally constructed.

We describe a real-time system for face and facial feature detection and tracking in continuous video. The core of this system consists of a set of novel facial feature detectors based on our previously proposed Information-Based Maximum Discrimination learning technique. These classiers are very fast and allow us to implement a fully automatic, real-time system for detection and tracking multiple faces. In addition to locking onto up to four target faces, this system locates and tracks nine facial features as they move under facial expression changes. 1 Introduction In this paper, we present in detail a fully automatic, person-independent, real-time system for detection and tracking multiple faces and nine facial features. We use Information-Based Maximum Discrimination classiers [1, 2] with a novel set of low-level image features to locate accurately and eÆciently nine facial features including non-rigid points as they move under facial expressions. 2 Tracking and Motion Analysi...

We present a probabilistic method to detect human faces using a mixture of factor analyzers. One characteristic of this mixture model is that it concurrently performs clustering and, within each cluster, local dimensionality reduction. A wide range of face images including ones in dierent poses, with dierent expressions and under dierent lighting conditions are used as the training set to capture the variations of human faces. In order to t the mixture model to the sample face images, the parameters are estimated using an EM algorithm. Experimental results show that faces in dierent poses, with dierent facial expressions, and under dierent lighting conditions are accurately detected by our method. 1. INTRODUCTION Images of human faces are central to intelligent human computer interaction. Much research is being done involving face images, including face recognition, face tracking, pose estimation, expression recognition and gesture recognition. However, most existing methods o...

Face recognition is one of the most interesting and challenging problems in computer vision. In the past, many facets of this problem have been rigorously investigated because of its importance for understanding our cognitive process and its usefulness in various applications. A great di#culty in face recognition is the separation of intrinsic facial characteristics from extrinsic image variations. Among the latter in 2D images is pose, illumination, and expression. Unfortunately, most past studies have provided variation-speci#c solutions that are not applicable to other types of variation. Performance has remained inferior to human ability and sub-optimal for practical use. This dissertation proposes a novel solution to one of these problems. We focus on processing head pose information in 2D images: analyzing, synthesizing, and identifying facial images with arbitrary pose. Successful handling of head pose variation is one of the key factors for realizing facial information processing systems in virtually any realistic and practical scenario. Our goal is twofold. One is to provide a simple and general framework whichmay be useful beyond the speci#c problem of head pose. The other is to improve the pose processing accuracy of previous studies by using this framework. xv We propose a localized two-stage linear system which is learned strictly from sample statistics and models shape and texture information separately. Instead of using variation-speci#c analytical knowledge of 3D rotation in Euclidean space, our solution utilizes a simple statistical learning framework whose applicability is not limited to the problem at hand. A wider range of head poses is covered byanumber of local linear models distributed over various poses, each of which realizes a continuous mapp...

WaveBase is a system for detecting features in a face image. It has a database of faces, each with a two-level hierarchical wavelet network. When a new face image is presented to the system for face detection, WaveBase searches its database for the "best face" -- the face whose first level wavelet network most closely matches the new face. It also determines an affine transformation to describe any difference in the orientation of the faces. By applying the affine transformation to the position of the features in the best face, approximate feature positions in the new face are found. Second level wavelet networks for each feature are then placed at these approximate positions, and allowed to move slightly to minimize their difference from the new face. This facilitates adjustments in addition to the affine transformation to account for slight differences in the geometry of the best head and the new head. The final position of the wavelet network is WaveBase's estimate of the feature positions. Experiments demonstrate the benefit of our hierarchical approach. Results compare favorably with existing techniques for feature localization.

A method of manifold representation for human faces with pose variations is proposed. Our model consists of mappings between 3D head angles and facial images separately represented in shape and texture, via sub-space models spanned by principal components (PCs). Explicit mappings to and from 3D head angles are used as processes of pose estimation and transformation, respectively. Generalization capability to unknown head poses enables our model to continuously cover pose parameter space, providing high approximation accuracy. The feasibility of this model is evaluated in a number of experiments. We also propose a novel pose-invariant face recognition system using our model as the entry format for a gallery of known persons. Experimental results with 3D facial models recorded by a Cyberware scanner show that our model provides a superior recognition performance against pose variations, and that texture synthesis process is carried out correctly. 1