Gaze Estimation Using Active Appearance Model Parameters Based on Regression Analysis
نویسندگان
چکیده
One of the most crucial techniques associated with computer vision is technology that deals with the automatic estimation of gaze orientation. In this paper, a method is proposed to estimate horizontal gaze orientation from a monocular camera image using parameters of an Active Appearance Model (AAM) based on linear regression or non-linear regression. The proposed method can estimate horizontal gaze orientation more precisely than a conventional method due to the simultaneous estimation of horizontal head pose and gaze orientation. Also, a non-linear method using Gaussian process regression is effective for cases in which a subject is wearing glasses, or the illumination is not sufficient. The validity of the proposed method was confirmed by experimental results.
منابع مشابه
Gaze Estimation Using Regression Analysis and AAMs Parameters Selected Based on Information Criterion
One of the most crucial techniques associated with Computer Vision is technology that deals with the automatic estimation of gaze orientation. In this paper, a method is proposed to estimate horizontal gaze orientation from a monocular camera image using the parameters of Active Appearance Models (AAM) selected based on several model selection methods. The proposed method can estimate horizonta...
متن کاملActive Appearance Models for Gaze Estimation
This thesis describes the implementation of and the experimentation with a gaze estimation system based on an approach called the Active Appearance Model (AAM). Active Appearance Models are generative models of a certain visual phenomenon. They can be used to model specific patterns of variability in shape and grey-level appearance, which in turn can be used directly in image interpretation. Ou...
متن کاملOn the Dimensionality of Deformable Face Models
Model-based face analysis is a general paradigm with applications that include face recognition, expression recognition, lipreading, head pose estimation, and gaze estimation. A face model is first constructed from a collection of training data, either 2D images or 3D range scans. The face model is then fit to the input image(s) and the model parameters used in whatever the application is. Most...
متن کاملGaze estimation using appearance models
A simple gaze estimation procedure based on appearance models is described. The method uses minimum norm linear regression to directly relate eye appearance to gaze direction. No explicit feature extraction is required. The method is fast, requiring only linear operations on the data. It is also accurate, in spite of using only a small amount of training data.
متن کاملPhD Thesis Enhanced Multiple Output Regression based on Canonical Correlation Analysis with Applications in Computer Vision
Modeling the functional relation between high dimensional signals is a common task in computer vision. Just like natural perception systems are able to relate and combine impressions from different senses (speech, facial expression, gestures, haptics, etc.), simultaneous processing of signals of different sources and establishing a functional relation between these sources is an important issue...
متن کامل