Use of depth and colour eigenfaces for face recognition

Tsalakanidou, Filareti; Tzovaras, Dimitrios; Strintzis, Michael G.

doi:10.1016/s0167-8655(02)00383-5

Cited by 137 publications

(86 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Tsalakanidou et al [93,92] propose a classic approach where shape and texture images are coded using PCA and their scores are fused at the decision level. Malassiotis and Strintzis [56] use an embedded hidden Markov model-based (EHMM) classifier which produces similarity scores and these scores are fused by a weighted sum rule.…”

Section: Feature Extraction and Matchingmentioning

confidence: 99%

3D Face Recognition: Technology and Applications

Gökberk

Salah

Alyüz

et al. 2009

Advances in Pattern Recognition

View full text Add to dashboard Cite

Section: Feature Extraction and Matchingmentioning

confidence: 99%

3D Face Recognition: Technology and Applications

Gökberk

Salah

Alyüz

et al. 2009

Advances in Pattern Recognition

View full text Add to dashboard Cite

“…As a real face is fully described by its 3D shape and its texture, it is reasonable to use both kind of data (geometry and color or intensity) to improve recognition reliability: this is the idea behind Multi-Modal or (3D+2D) face recognition. The work by Tsalakanidou et al (Tsalakanidou et al, 2003) is based on PCA to compare both probe's range image and intensity/color image to the gallery, Papatheodorou and Rueckert (Papatheodorou & Rueckert, 2004) presented a 4D registration method based on Iterative Closest Point (ICP), augmented with texture data. Bronstein et al (Bronstein et al, 2003) propose a multi-modal 3D + 2D recognition using eigen decomposition of flattened textures and canonical images.…”

Section: Related Workmentioning

confidence: 99%

3D Face Recognition in a Ambient Intelligence Environment Scenario

Abate¹,

Ricciardi²,

Sabatino³

2007

Face Recognition

View full text Add to dashboard Cite

“…In (19) , point signatures have also been proposed for face identification. In (20) (21) (23) , a face recognition technique is developed based on 3D shape and texture information. In (26) , we had performed 3D face recognition followed by extraction of facial contour lines at a series of different depth values.…”

Section: Introductionmentioning

confidence: 99%

3D Face Recognition Using Parallel Pyramid Neural Networks

2006

View full text Add to dashboard Cite

Face recognition plays an important role in many applications such as building/store access control, suspect identification and surveillance. However there are several problems that make automatic face recognition to be a very difficult task. The input face image of a person presented to a face recognition system is not usually taken under similar conditions as the face image of the same person in the database. Therefore, it is important that an automatic face recognition system is able to cope with numerous possible variations among different images of the same face. In order to solve this problem, a pose, illumination condition and face expressions invariant 3D face recognition system are proposed in this paper.Original facial images were obtained by RangeFinder Scansion 3D [Danae-R], example images in database are illustrated in Fig. 1. We first compensate for the poses of 3D original facial images using feature points and geometrical measurement. To extract the features of original shape and texture spaces, PCA and LDA algorithms are used to obtain low dimensional shape and texture features, respectively. The low dimensional shape and texture features of each image are then combined to one vector for recognition. Fig. 1. Example images in databaseIn our recognition system, a parallel pyramid NNs is proposed. NN has been widely used in face recognition field. However, if there are a few samples of each registrant in the training set, the generalization of conventional NN will be poor in the training process. In order to solve this problem, the pyramid NN is adopted since it is a nonlinear alternative with better generalization. In addition, if there are a great number of registrants in the training set, some problems are encountered such as the lower recognition rate and very slow convergence speed in training. In this study, the registrants in the training set are divided into several small-scale parallel NNs and they are combined to obtained the recognition result. Furthermore, we proposed a new autonomic control method for parallel NNs on a cross-validation set. The structure of the proposed parallel neural networks is illustrated in Fig. 2.Experimental results using 300 face images corresponding to 60 individuals of varying illumination, pose and facial expression, show that 1) the integrated shape and texture features carry the most discriminating information followed in order by using only textures images and shape images only and 2) our proposed system performs the best among the mosaic picture method and Eigenfaces method. In particular, proposed system achieves 99.17% recognition rate using only 120 features. Performance comparisons of various face features and classifiers is presented in Table 1. Non-memberIn this paper, we propose a method of 3D face recognition using parallel pyramid neural networks (NNs). We first compensate for the poses of 3D original facial images using feature points and geometrical measurement. Then, the shape and texture images are extracted from compensated 3D images, re...

show abstract

Use of depth and colour eigenfaces for face recognition

Cited by 137 publications

References 10 publications

3D Face Recognition: Technology and Applications

3D Face Recognition: Technology and Applications

3D Face Recognition in a Ambient Intelligence Environment Scenario

3D Face Recognition Using Parallel Pyramid Neural Networks

Contact Info

Product

Resources

About