Mircea C. Ionita

Separating Directional Lighting Variability in Statistical Face Modelling Based on Texture Space Decomposition

In this paper we propose a simple method for decomposing the linear texture space of a facial appearance model into two linear subspaces, one for inter-individual variability and another for variations caused by directional changes of the lighting conditions. The approach used is to create one linear subspace from individuals with uniform illumination conditions and then filter a set of images with various directional lighting conditions by projecting corresponding textures on the previously built space; the residues are further used to build a second subspace for directional lighting. The resulted subspaces are orthogonal, so the overall texture model can be obtained by a simple concatenation of the two subspaces. The main advantage of this representation is that two sets of parameters are used to control inter-individual variation and separately intra-individual variation due to changes in illumination conditions.

Statistical models of appearance for eye tracking and eye-blink detection and measurement

A statistical active appearance model (AAM) is developed to track and detect eye blinking. The model has been designed to be robust to variations of head pose or gaze. In particular we analyze and determine the model parameters which encode the variations caused by blinking. This global model is further extended using a series of sub-models to enable independent modeling and tracking of the two eye regions. Several methods to enable measurement and detection of eye-blink are proposed and evaluated. The results of various tests on different image databases are presented to validate each model.

Next Generation Face Tracking Technology Using AAM Techniques

This paper realizes an investigation of the active appearance model (AAM) techniques, with particular emphasis on problems related to the implementation of AAM in realtime face tracking applications for embedded systems. The paper includes (a) new training and model construction approaches for AAMs with increased robustness to head pose variations and changes in the illumination conditions, and (b) a corresponding AAM-based face tracking scheme.

Benefits of Using Decorrelated Color Information for Face Segmentation/Tracking

We analyze in this paper the benefits that can be derived from employing color image alignment techniques in the context of face segmentation or tracking based on texture (defined as the patch of intensities) template matching. By making full use of the decorrelated color information, improvements on the accuracy of the segmentation are demonstrated. This is intended to enhance the face segmentation algorithm by increasing its robustness to differences in images caused by various image acquisition devices or settings or by variations in the ambient illumination conditions.

Smart cameras: 2D affine models for determining subject facial expressions

In this paper we consider how next generation of smart cameras will move beyond face detection and tracking to develop more useful analysis tools for determining the expressions and emotions of the subjects in images. An initial proof-of-concept using an extension of an Active Appearance Model (AAM) to measure facial parameters and a set of classifiers to determine facial states are described. Preliminary findings are encouraging and suggest that the use of the component-based AAM models and SVM classifiers will enable new applications and opportunities in smart cameras.

On Color Texture Normalization for Active Appearance Models

The extension of the standard grayscale active appearance model (AAM) techniques to color images is investigated. Prior work in this field has mainly focused on RGB color models which did not demonstrate noticeable benefits over grayscale models from the point of view of convergence accuracy. We improve on previous work by normalizing the color texture vector separately for intensity and chromaticity components. Where an appropriate color space is chosen, we demonstrate improvements in convergence accuracy as well as image synthesization quality for AAMs. Optimal results are achieved when a color space in which the image channels are strongly decorrelated is chosen. Our best results are achieved using the I1I2I3 color space, originally proposed by Ohta.

A Statistical Modeling based System for Blink Detection in Digital Cameras

In this paper we describe a system which uses statistical eye modeling to track the eyes and detect eyes blinking. The main challenge was to design an eye model robust to small variations of head pose and to extract the parameters of the model which encode the variation caused by blinking.

Enhanced real-time face models from stereo imaging for gaming applications

Techniques for improved 2D active appearance face models are presented. When these are applied to stereoscopic image pairs we show that sufficient information on image depth is obtained to generate an approximate 3D face model. Two techniques are investigated, the first based on 2D+3D AAMs and the second using methods based on thin plate splines. The resulting 3D models can offer a practical real-time face model which is suitable for a range of applications in computer gaming. Due to the compact nature of AAMs these are also very suitable for use in embedded devices such as gaming peripherals!!

Smart cameras: 2D affine models for robust person recognition in consumer images

In this paper we consider how next generation smart cameras will move beyond face detection and tracking to develop more useful analysis tools for determining the expression and emotions of the subjects in images. An initial proof-of-concept using an Active Appearance Model (AAM) to perform face recognition which is robust to pose and illumination variations is presented. Our initial results are very encouraging and suggest that the incorporation of a specialized AAM within next generation digital cameras will enable more accurate management and sorting of personal image collections based on the actual persons in the images.

Appearance capturing and animation techniques for realistic, true-identity gaming avatars

Making use of a class of 2D statistical face models known as active appearance models (AAMs), we investigate a technique for generating real-time animated avatars suitable for video gaming. We consider here a number of improvements of these models in order to make them more robust to varying environmental conditions, as well as to reduce the computational load such that to become adapted for real-time implementations.