3D Face Pose Estimation Research Articles

Cumulative attribute space regression for head pose estimation and color constancy

Two-stage Cumulative Attribute (CA) regression has been found effective in regression problems of computer vision such as facial age and crowd density estimation. The first stage regression maps input features to cumulative attributes that encode correlations between target values. The previous works have dealt with single output regression. In this work, we propose cumulative attribute spaces for 2- and 3-output (multivariate) regression. We show how the original CA space can be generalized to multiple output by the Cartesian product (CartCA). However, for target spaces with more than two outputs the CartCA becomes computationally infeasible and therefore we propose an approximate solution - multi-view CA (MvCA) - where CartCA is applied to output pairs. We experimentally verify improved performance of the CartCA and MvCA spaces in 2D and 3D face pose estimation and three-output (RGB) illuminant estimation for color constancy.

Discriminant sparse label-sensitive embedding: Application to image-based face pose estimation

In this letter, the authors propose a new embedding scheme for image-based continuous face pose estimation. The main contributions are as follows. First, it is shown that the concept of label-sensitive Locality Preserving Projections, proposed for age estimation, can be used for model-less face pose estimation. Second, the authors propose a linear embedding by exploiting the connections between facial features and pose labels via a sparse coding scheme. The resulting technique is called Sparse Label sensitive Locality Preserving Projections (Sp-LsLPP). Third, for enhancing the discrimination between poses, the projections obtained by Sp-LsLPP are fed to a Discriminant Embedding that exploits the continuous labels. The resulting framework has less parameters compared to related works. It has been applied to the problem of model-less face yaw angle estimation (person independent 3D face pose estimation). It was tested on three databases: FacePix, Taiwan, and Columbia. It was conveniently compared with other linear and non-linear techniques. The experimental results confirm that the proposed framework can outperform, in general, the existing ones.

3D face pose estimation by a robust real time tracking of facial features

In this paper, we present a new 3D head pose estimating approach based on real-time facial feature tracking and facial feature recovering method which copes with the surrounding light variation and various occlusion. The major facial features are obtained by Haar-like feature detection along with AdaBoost learning from an input video image. The detected facial features are robustly tracked by optical flow with a template matching scheme which continuously compensates for losing track of the initially detected features in a sequence of input images. The head pose of an input face image can be obtained by evaluating 3D information of facial features from the detected 2D eye-points, nose and lip. From the experiments, the proposed method shows effectiveness in tracking and recovering facial features and produces reliable result in head pose estimation.

A central profile-based 3D face pose estimation

In this study, we develop a central profile-based 3D face pose estimation algorithm. The central profile is a unique curve on a 3D face surface that starts from forehead center, goes down through nose ridge, nose tip, mouth center, and ends at a chin tip. The points on the central profile are co-planar and belong to a symmetry plane that separates human face into two identical parts. The central profile is protrusive and has a certain length. Most importantly, the normal vectors of the central profile points are parallel to the symmetry plane. Based on the properties of the central profile, Hough transform is employed to determine the symmetry plane by invoking a voting procedure. An objective function is introduced in the parameter space to quantify the vote importance for face profile points and map the central profile to an accumulator cell with the maximal value. Subsequently, a nose model matching algorithm is used to detect nose tip on the central profile. A pitch angle estimation algorithm is also proposed. The pose estimation experiments completed for a synthetic 3D face model and the FRGC v2.0 3D database demonstrate the effectiveness of the proposed pose estimation algorithm. The obtained central profile detection rate is 99.9%, and the nose tip detection rate has reached 98.16% with error not larger than 10mm.

3D Face pose estimation and tracking from a monocular camera

Abstract In this paper, we describe a new approach for estimating and tracking three-dimensional (3D) pose of a human face from the face images obtained from a single monocular view with full perspective projection. We assume that the shape of a 3D face can be approximated by an ellipse and that the aspect ratio of 3D face ellipse is given. Given a monocular image of a face, we first perform an ellipse detection to locate the face in the image and the 3D position and orientation of the face are then estimated from the detected image face. The face detection is greatly facilitated by exploring the physiological properties of eyes under a special IR illumination and some geometric constraints. The detected initial face ellipse is then tracked in subsequent frames, allowing to track 3D face pose from frame to frame. Compared with the existing feature-based approaches for face pose estimation, our approach is more robust, since ellipse can be more reliably and robustly detected and tracked. Experimental study using a large number of synthetic and real images demonstrates the accuracy and robustness of the proposed approach.