Antonio Rama

Mixed 2D-3D Information for Pose Estimation and Face Recognition

Face recognition based on 3D techniques is a promising approach since it takes advantage of the additional information provided by depth which makes the whole approach more robust against illumination and pose variations. However, these 3D approaches require the cooperation of the person to acquire accurate 3D data; thus, they are not appropriated for some applications such as video surveillance or restricted area access points where only a 2D face image is disposable. In this paper, a novel approach is presented which takes advantage of 3D data in the training stage but only requires 2D data in the recognition stage. The proposed method can be used for both pose estimation and face recognition. Moreover, the estimation of the pose can be used as side information to improve the performance of the face recognition stage. Experiments have been carried out on the public UPC face database which is composed of a total of 621 face images of several persons taken from different views and illuminations

More robust face recognition by considering occlusion information

This paper addresses one of the main challenges of face recognition (FR): facial occlusions. Currently, the human brain is the most robust known FR approach towards partially occluded faces. Nevertheless, it is still not clear if humans recognize faces using a holistic or a component-based strategy, or even a combination of both. In this paper, three different approaches based on principal component analysis (PCA) are analyzed. The first one, a holistic approach, is the well-known eigenface approach. The second one, a component-based method, is a variation of the eigenfeatures approach, and finally, the third one, a near-holistic method, is an extension of the lophoscopic principal component analysis (LPCA). So the main contributions of this paper are: The three different strategies are compared and analyzed for identifying partially occluded faces and furthermore it explores how a priori knowledge about present occlusions can be used to improve the recognition performance.

Face Recognition using a Fast Model Synthesis from a Profile and a Frontal View

In our previous work we presented a new 2D-3D mixed face recognition scheme called Partial Principal Component Analysis (P2CA). The main contribution of P2CA is that it uses 3D data in the training stage but it accepts either 2D or 3D information in the recognition stage. We think that 2D-3D mixed approaches are the next step in face recognition research since most of surveillance or access control applications only dispose of a single camera which is used to acquire a single 2D texture image. Nevertheless, one of the main problems of our previous work was the enrollment of new persons in the database (gallery set) since a total of five different pictures are needed for getting the 180deg texture maps (manual morphing). Thus, this work is focused on the automatic and fast creation of those 180deg texture maps from only two images (frontal and profile views). Preliminary results show that there is not a significant degradation of the recognition accuracy when using this automatically and synthetically created gallery set instead of the one created by morphing the five views manually.

Aligned texture map creation for pose invariant face recognition

In last years, Face recognition based on 3D techniques is an emergent technology which has demonstrated better results than conventional 2D approaches. Using texture (180° multi-view image) and depth maps is supposed to increase the robustness towards the two main challenges in Face Recognition: Pose and illumination. Nevertheless, 3D data should be acquired under highly controlled conditions and in most cases depends on the collaboration of the subject to be recognized. Thus, in applications such as surveillance or control access points, this kind of 3D data may not be available during the recognition process. This leads to a new paradigm using some mixed 2D-3D face recognition systems where 3D data is used in the training but either 2D or 3D information can be used in the recognition depending on the scenario. Following this concept, where only part of the information (partial concept) is used in the recognition, a novel method is presented in this work. This has been called Partial Principal Component Analysis (P2CA) since they fuse the Partial concept with the fundamentals of the well known PCA algorithm. This strategy has been proven to be very robust in pose variation scenarios showing that the 3D training process retains all the spatial information of the face while the 2D picture effectively recovers the face information from the available data. Furthermore, in this work, a novel approach for the automatic creation of 180° aligned cylindrical projected face images using nine different views is presented. These face images are created by using a cylindrical approximation for the real object surface. The alignment is done by applying first a global 2D affine transformation of the image, and afterward a local transformation of the desired face features using a triangle mesh. This local alignment allows a closer look to the feature properties and not the differences. Finally, these aligned face images are used for training a pose invariant face recognition approach (P2CA).

Fast and Robust Graphic Character Verification System for TV Sets

In this paper, we present a flexible system to verify any kind of graphic TV character or symbol in television charts. The proposed approach is very robust towards translations, rotations, changes in scale and illumination variations of the graphic characters. Another important attribute of the proposed system is its speed. The system should be very fast since the characters have to be verified in a real production line of TV sets. The proposed system is composed of fast and accurate preprocessing modules in order to normalize the characters, and a PCA- based verification method. Results show a very high performance with verification rates over the 99.8%.

Partial LDA vs Partial PCA

Recently, 3D face recognition algorithms have outperformed 2D conventional approaches by adding depth data to the problem. However, independently of the nature (2D or 3D) of the approach, the majority of them required the same data format in the test stage than the data used for training the system. This issue represents the main drawback of 3D face research since 3D data should be acquired under highly controlled conditions and in most cases require the collaboration of the subject to be recognized. Thus, in real world applications (control access points or surveillance) this kind of 3D data may not be available during the recognition process. This leads to a new paradigm using some mixed 2D-3D face recognition systems where 3D data is used in the training but either 2D or 3D information can be used in the recognition depending on the scenario. Following this new concept, partial linear discriminant analysis (PLDA) is presented in this paper. Preliminary results have shown an improvement with respect to the partial PCA approach

Cartoon Detection Using Integral

With the growth of digital television TV program classification has become a major research topic. Recent classification techniques have reported acceptable results for specific genre detection. Cartoons is one of these genres which has deceived some attention because of its importance in push scenarios where parents want to control their children s access to television. In this paper a flexible scheme based on a non-linear classifier called fuzzy integral is presented. This operator is supposed not only to classify but also to give a relevance measure to all the features involved in the classification. Preliminary results using this operator for cartoon detection are presented and compared with other well known statistical clarification methods like PCA, IDA or K-NN.

Using partial information for face recognition and pose estimation

The main achievement of this work is the development of a new face recognition approach called partial principal component analysis (P/sup 2/CA), which exploits the novel concept of using only partial information for the recognition stage. This approach uses 3D data in the training stage but it permits to use either 2D or 3D data in the recognition stage, making the whole system more flexible. Preliminary experiments carried out on a multi-view face database composed of 18 individuals have shown robustness against big pose variations obtaining higher recognition rates than the conventional PCA method. Moreover, the P/sup 2/CA method can estimate the pose of the face under different illuminations with accuracy of the 96.15% when classifying the face images in 0/spl deg/, /spl plusmn/30/spl deg/, /spl plusmn/45/spl deg/, /spl plusmn/60/spl deg/ and /spl plusmn/90/spl deg/ views.

2D-3D Pose Invariant Face Recognition System for Multimedia Applications

Automatic Face recognition of people is a challenging problem which has received much attention during the recent years due to its potential multimedia applications in different fields such as 3D videoconference, security applications or video indexing. However, there is no technique that provides a robust solution to all situations and different applications, yet. Face recognition includes a set of challenges like expression variations, occlusions of facial parts, similar identities, resolution of the acquired images, aging of the subjects and many others. Among all these challenges, most of the face recognition techniques have evolved in order to overcome two main problems: illumination and pose variation. Either of these influences can cause serious performance degradation in a 2D face recognition system.Some of the new face recognition strategies tend to overcome both research topics from a 3D perspective. The 3D data points corresponding to the surface of the face may be acquired using different alternatives: A multi camera system (stereoscopy), structured light, range cameras or 3D laser and scanner devices The main advantage of using 3D data is that geometry information does not depend on pose and illumination and therefore the representation of the object does not change with these parameters, making the whole system more robust. However, the main drawback of the majority of 3D face recognition approaches is that they need all the elements of the system to be well calibrated and synchronized to acquire accurate 3D data (texture and depth maps). Moreover, most of them also require the cooperation or collaboration of the subject during a certain period of time. All these requirements can be available during the training stage of many applications. When enrolling a new person in the database, it could be performed off-line, with the help o human interaction and with the cooperation of the subject to be enrolled. On the contrary, the previous conditions are not always available during the test stage. The recognition will be in most of the cases in a semicontrolled or uncontrolled scenario, where the only input of the system will probably consists of a 2D intensity image acquired from a single camera. This leads to a new paradigm where 2D-3D mixed face recognition approaches are used. The idea behind this kind of approaches is that these take profit of the 3D data during the training stage but then they can use either 3D data (when available) or 2D data during the recognition stage. Belonging to this category, some of 2D statistical approaches like Eigenfaces of Fisherfaces have been extended to fit in this new paradigm leading to the Partial Principal Component Analysis (P2CA) approach. This algorithm intends to cope with big pose variations (±90 ∘) by using 180∘ cylindrical texture maps for training the system but then only images acquired from a single, normal camera are used for the recognition. These training images provide pose information from different views (2.5D data). Nevertheless they can also be extended to a complete 3D multimodal system where depth and texture information is used. This chapter is structured as follows: First, a brief overview of the state-of-the-art in face recognition is introduced. The most relevant methods are grouped by multimedia scenarios and concrete applications. Afterwards, novel 2D-3D mixed face recognition approaches will be introduced.

Cascade scheme face detection using a non-liniar classifier

In this paper, the non-linear fuzzy integral operator is proposed for combining different sets of Haar features for face detection. The proposed method presents a lower false detection rate than the State-of-the-art AdaBoost face detector by a similar true acceptance rate and using the same optimal set of features. Furthermore, this novel face detector seems to have a better generalization capability than the AdaBoost method. Experimental results show a positive face detection rate larger than 92% and a false detection rate of 0.1% when using a four stage cascade scheme.