... Traditional database techniques have been adequate for many applications involving alphanumer... more ... Traditional database techniques have been adequate for many applications involving alphanumeric records, which ... 4] as well as fuzzy logic [5]. In some applications such as ... Multimedia Information systems, volume 1508 of Lecture Notes in Computer Science, International, pp. ...
SMC'98 Conference Proceedings. 1998 IEEE International Conference on Systems, Man, and Cybernetics (Cat. No.98CH36218), 1998
Increasing the search speed for matching range and domain blocs is the main challenge facing frac... more Increasing the search speed for matching range and domain blocs is the main challenge facing fractal-based images compression. One way to remedy at this problem is to classify image blocs into categories and only search among domain blocs which are in the same category as the target range bloc. Since image blocs with a simple edge are a very important
International Symposium on Control, Communications and Signal Processing, 2012
This paper addresses the issue of Gender Classification from 3D facial images. While most of prev... more This paper addresses the issue of Gender Classification from 3D facial images. While most of previous work in the literature focuses on either 2D facial images, here, we study the use of 3D facial shape for automatic gender classification. After a preprocessing step to extract the facial masks from triangular meshes obtained using laser range scanners, we approximate the facial
ABSTRACT In this paper we present a fully automatic approach for identity-independent facial expr... more ABSTRACT In this paper we present a fully automatic approach for identity-independent facial expression recognition from 3D video sequences. Towards that goal, we propose a novel approach to extract a scalar field that represents the deformations between faces conveying different expressions. We extract relevant features from this deformation field using LDA and then train a dynamic model on these features using HMM. Experiments conducted on BU-4DFE dataset following state-of-the-art settings show the effectiveness of the proposed approach.
In this paper, we present an automatic approach for facial expression recognition from 3-D video ... more In this paper, we present an automatic approach for facial expression recognition from 3-D video sequences. In the proposed solution, the 3-D faces are represented by collections of radial curves and a Riemannian shape analysis is applied to effectively quantify the deformations induced by the facial expressions in a given subsequence of 3-D frames. This is obtained from the dense scalar field, which denotes the shooting directions of the geodesic paths constructed between pairs of corresponding radial curves of two faces. As the resulting dense scalar fields show a high dimensionality, Linear Discriminant Analysis (LDA) transformation is applied to the dense feature space. Two methods are then used for classification: 1) 3-D motion extraction with temporal Hidden Markov model (HMM) and 2) mean deformation capturing with random forest. While a dynamic HMM on the features is trained in the first approach, the second one computes mean deformations under a window and applies multiclass random forest. Both of the proposed classification schemes on the scalar fields showed comparable results and outperformed earlier studies on facial expression recognition from 3-D video sequences.
Three-Dimensional Imaging, Interaction, and Measurement, 2011
We present a novel method for 3D-shape matching using Bag-of-Feature techniques (BoF). The method... more We present a novel method for 3D-shape matching using Bag-of-Feature techniques (BoF). The method starts by selecting and then describing a set of points from the 3D-object. Such descriptors have the advantage of being invariant to different transformations that a shape ...
In the present paper, a general affine motion estimation algorithm using the Fourier descriptors ... more In the present paper, a general affine motion estimation algorithm using the Fourier descriptors is proposed. It is able to estimate, in addition to the translation, scaling and rotation, the stretching parameters or equivalently, the scale, the shift, and the four coefficients of the general affine matrix. A modified Claire test sequence will serve as a useful test to evaluate
Three-Dimensional face recognition is a challenging task with a large number of proposed solution... more Three-Dimensional face recognition is a challenging task with a large number of proposed solutions [1, 2]. With variations in pose and expression the identification of a face scan based on 3D geometry is difficult. To improve on this task and to evaluate existing face matching methods large sets of 3D faces were constructed, such as the FRGC [3], BU-3DFE [4],
ABSTRACT Although it is valuable information that human faces are approximately symmetric, in the... more ABSTRACT Although it is valuable information that human faces are approximately symmetric, in the literature of facial attributes recognition, little consideration has been given to the relationship between gender, age, ethnicity, etc. and facial asymmetry. In this paper we present a new approach based on bilateral facial asymmetry for gender classification. For that purpose, we propose to first capture the facial asymmetry by using Deformation Scalar Field (DSF) applied on each 3D face, then train such representations (DSFs) with several classifiers, including Random Forest, Adaboost and SVM after PCAbased feature space transformation. Experiments conducted on FRGCv2 dataset showed that a significant relationship exists between gender and facial symmetry when achieving a 90.99% correct classification rate for the 466 earliest scans of subjects (mainly neutral) and 88.12% on the whole FRGCv2 dataset (including facial expressions).
ABSTRACT The 3D face recognition literature has many papers that represent facial shapes as colle... more ABSTRACT The 3D face recognition literature has many papers that represent facial shapes as collections of curves of different kinds (level-curves, iso-level curves, radial curves, profiles, geodesic polarization, iso-depth lines, iso-stripes, etc.). In contrast with the holistic approaches, the approaches that match faces based on whole surfaces, the curve-based parametrization allows local analysis of facial shapes. This, in turn, facilitates handling of pose variations (probe image may correspond to a part of the face) or missing data (probe image is altered by occlusions. An important question is: Does the use of full set of curves leads to better performances? Among all facial curves, are there ones that are more relevant than others for the recognition task? We explicitly address these questions in this paper. We represent facial surfaces by collections of radial curves and iso-level curves, such that shapes of corresponding curves are compared using a Riemmannian framework, select the most discriminative curves (geometric features) using boosting. The experiment involving FRGCv2 dataset demonstrates the effectiveness of this feature selection by achieving 98.02% as rank-1 recognition rate. This selection also results in a more compact signature which significantly reduces the computational cost and the storage requirements for the face recognition system.
ABSTRACT This paper addresses the issue of Gender Classification from 3D facial images. While mos... more ABSTRACT This paper addresses the issue of Gender Classification from 3D facial images. While most of previous work in the literature focuses on either 2D facial images, here, we study the use of 3D facial shape for automatic gender classification. After a preprocessing step to extract the facial masks from triangular meshes obtained using laser range scanners, we approximate the facial surfaces by collections of radial and iso-level curves. Once the curves are extracted, we aim at studying their shape using existant shape analysis framework which allows to compute similarities between a candidate face and Male and Female templates. We expect that the shape of certain curves are similar within Male/Female classes and different when moving from one class to another. For classification, we perfom three Machine Learning algorithms (Adaboost, Neural Network, and SVM). Overall, Adaboost was superior in classification performance (84.98% as classification rate) on a subset of FRGCv2 dataset including the first (neutral and non-neutral) scans of different subjects. Our results indicate also that (i) the most relevant iso-level curves cover the central stripe of the face, and (ii) the most relevant radial curves are located on the upper part of the face.
... Traditional database techniques have been adequate for many applications involving alphanumer... more ... Traditional database techniques have been adequate for many applications involving alphanumeric records, which ... 4] as well as fuzzy logic [5]. In some applications such as ... Multimedia Information systems, volume 1508 of Lecture Notes in Computer Science, International, pp. ...
SMC'98 Conference Proceedings. 1998 IEEE International Conference on Systems, Man, and Cybernetics (Cat. No.98CH36218), 1998
Increasing the search speed for matching range and domain blocs is the main challenge facing frac... more Increasing the search speed for matching range and domain blocs is the main challenge facing fractal-based images compression. One way to remedy at this problem is to classify image blocs into categories and only search among domain blocs which are in the same category as the target range bloc. Since image blocs with a simple edge are a very important
International Symposium on Control, Communications and Signal Processing, 2012
This paper addresses the issue of Gender Classification from 3D facial images. While most of prev... more This paper addresses the issue of Gender Classification from 3D facial images. While most of previous work in the literature focuses on either 2D facial images, here, we study the use of 3D facial shape for automatic gender classification. After a preprocessing step to extract the facial masks from triangular meshes obtained using laser range scanners, we approximate the facial
ABSTRACT In this paper we present a fully automatic approach for identity-independent facial expr... more ABSTRACT In this paper we present a fully automatic approach for identity-independent facial expression recognition from 3D video sequences. Towards that goal, we propose a novel approach to extract a scalar field that represents the deformations between faces conveying different expressions. We extract relevant features from this deformation field using LDA and then train a dynamic model on these features using HMM. Experiments conducted on BU-4DFE dataset following state-of-the-art settings show the effectiveness of the proposed approach.
In this paper, we present an automatic approach for facial expression recognition from 3-D video ... more In this paper, we present an automatic approach for facial expression recognition from 3-D video sequences. In the proposed solution, the 3-D faces are represented by collections of radial curves and a Riemannian shape analysis is applied to effectively quantify the deformations induced by the facial expressions in a given subsequence of 3-D frames. This is obtained from the dense scalar field, which denotes the shooting directions of the geodesic paths constructed between pairs of corresponding radial curves of two faces. As the resulting dense scalar fields show a high dimensionality, Linear Discriminant Analysis (LDA) transformation is applied to the dense feature space. Two methods are then used for classification: 1) 3-D motion extraction with temporal Hidden Markov model (HMM) and 2) mean deformation capturing with random forest. While a dynamic HMM on the features is trained in the first approach, the second one computes mean deformations under a window and applies multiclass random forest. Both of the proposed classification schemes on the scalar fields showed comparable results and outperformed earlier studies on facial expression recognition from 3-D video sequences.
Three-Dimensional Imaging, Interaction, and Measurement, 2011
We present a novel method for 3D-shape matching using Bag-of-Feature techniques (BoF). The method... more We present a novel method for 3D-shape matching using Bag-of-Feature techniques (BoF). The method starts by selecting and then describing a set of points from the 3D-object. Such descriptors have the advantage of being invariant to different transformations that a shape ...
In the present paper, a general affine motion estimation algorithm using the Fourier descriptors ... more In the present paper, a general affine motion estimation algorithm using the Fourier descriptors is proposed. It is able to estimate, in addition to the translation, scaling and rotation, the stretching parameters or equivalently, the scale, the shift, and the four coefficients of the general affine matrix. A modified Claire test sequence will serve as a useful test to evaluate
Three-Dimensional face recognition is a challenging task with a large number of proposed solution... more Three-Dimensional face recognition is a challenging task with a large number of proposed solutions [1, 2]. With variations in pose and expression the identification of a face scan based on 3D geometry is difficult. To improve on this task and to evaluate existing face matching methods large sets of 3D faces were constructed, such as the FRGC [3], BU-3DFE [4],
ABSTRACT Although it is valuable information that human faces are approximately symmetric, in the... more ABSTRACT Although it is valuable information that human faces are approximately symmetric, in the literature of facial attributes recognition, little consideration has been given to the relationship between gender, age, ethnicity, etc. and facial asymmetry. In this paper we present a new approach based on bilateral facial asymmetry for gender classification. For that purpose, we propose to first capture the facial asymmetry by using Deformation Scalar Field (DSF) applied on each 3D face, then train such representations (DSFs) with several classifiers, including Random Forest, Adaboost and SVM after PCAbased feature space transformation. Experiments conducted on FRGCv2 dataset showed that a significant relationship exists between gender and facial symmetry when achieving a 90.99% correct classification rate for the 466 earliest scans of subjects (mainly neutral) and 88.12% on the whole FRGCv2 dataset (including facial expressions).
ABSTRACT The 3D face recognition literature has many papers that represent facial shapes as colle... more ABSTRACT The 3D face recognition literature has many papers that represent facial shapes as collections of curves of different kinds (level-curves, iso-level curves, radial curves, profiles, geodesic polarization, iso-depth lines, iso-stripes, etc.). In contrast with the holistic approaches, the approaches that match faces based on whole surfaces, the curve-based parametrization allows local analysis of facial shapes. This, in turn, facilitates handling of pose variations (probe image may correspond to a part of the face) or missing data (probe image is altered by occlusions. An important question is: Does the use of full set of curves leads to better performances? Among all facial curves, are there ones that are more relevant than others for the recognition task? We explicitly address these questions in this paper. We represent facial surfaces by collections of radial curves and iso-level curves, such that shapes of corresponding curves are compared using a Riemmannian framework, select the most discriminative curves (geometric features) using boosting. The experiment involving FRGCv2 dataset demonstrates the effectiveness of this feature selection by achieving 98.02% as rank-1 recognition rate. This selection also results in a more compact signature which significantly reduces the computational cost and the storage requirements for the face recognition system.
ABSTRACT This paper addresses the issue of Gender Classification from 3D facial images. While mos... more ABSTRACT This paper addresses the issue of Gender Classification from 3D facial images. While most of previous work in the literature focuses on either 2D facial images, here, we study the use of 3D facial shape for automatic gender classification. After a preprocessing step to extract the facial masks from triangular meshes obtained using laser range scanners, we approximate the facial surfaces by collections of radial and iso-level curves. Once the curves are extracted, we aim at studying their shape using existant shape analysis framework which allows to compute similarities between a candidate face and Male and Female templates. We expect that the shape of certain curves are similar within Male/Female classes and different when moving from one class to another. For classification, we perfom three Machine Learning algorithms (Adaboost, Neural Network, and SVM). Overall, Adaboost was superior in classification performance (84.98% as classification rate) on a subset of FRGCv2 dataset including the first (neutral and non-neutral) scans of different subjects. Our results indicate also that (i) the most relevant iso-level curves cover the central stripe of the face, and (ii) the most relevant radial curves are located on the upper part of the face.
Uploads
Papers by Mohamed Daoudi