S. J. McKenna, S. Gong, R. P. Wurtz, J. Tanner, D. Banin "Tracking Facial Feature Points with Gabor Wavelets and Shape Models" 1st Int. Conf. on Audio- and Videobased Biometric Person Authentication, Lecture Notes in Computer Science 1997.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....of variability in shape, texture, pose, and imaging conditions. Detecting faces in images has received much attention. A comprehensive survey on face detection methods can be found in [1] A huge research effort has been devoted to detecting and tracking of facial features in 2D and 3D (e.g. [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]) Recovering the face position and the facial expression automatically from a video is a difficult problem. This problem can be made easier by using markers on the face such as heavy makeup or a set of colored dots stuck onto the face [14] Although the use of markers reduces the difficulty of ....

S.J. McKenna, S. Gong, R.P. Wurtz, J. Tanner, and D. Banin, "Tracking facial feature points with Gabor wavelets and shape models," in Int. Conference on Audio- and Video-based Biometric Person Authentication, 1997.

....the next generation of visual telecommunication due to its flexibility and efficiency. Model based image coding schemes can be divided into two types. One uses explicit face model to analyze moving facial images (e.g, see [8] In the second type, there is no explicit face model. For example, in [9], deformable 2D shapes are represented using Point Distribution Models. This paper describes how we use an active model to adapt a deformable wireframe model to the face in each input image, i.e. to track the face through an image sequence. Central to our approach is the decoupling of global head ....

S. McKenna, S. Gong, R. Wurtz, J. Tanner, and D. Banin. Tracking facial feature points with Gabor wavelets and shape models. In Int. Conference on Audio- and Video-based Biometric Person Authentication, 1997.

....computer vision and signal processing communities, where ability to track motion is useful in applications such as face based biometrics person authentication, facial expression analysis, animation or teleconferencing. Since snakes appear, a lot of work has been done to recover non rigid motion [1, 4, 6, 5, 9]. While snakes enforce smoothness in their shapes, they do not take into account restrictions for a particular non rigid motion and they are sensitive to occlusion and noise. Other trackers have emerged to solve these problems, such as Active Contours [1] all techniques relaying on template ....

....non rigid motion and they are sensitive to occlusion and noise. Other trackers have emerged to solve these problems, such as Active Contours [1] all techniques relaying on template matching in image or Eigenspace [4, 12, 10] the ones using optical flow [2] deformable templates, feature trackers [9] and Active Shape Models [6] Active contours [1] are not well suited for tracking complex objects without contextual information or special signals. All optical flow techniques [2] have reported good results for recognition purposes, but are not valid ones for accurate tracking, since the optical ....

[Article contains additional citation context not shown here]

S. McKenna, S. Gong, R. Wurtz, J. Tanner, and D. Banin. Tracking facial feature points with gabor wavelets and shape models. In Audio-Video Based Biome. Person Auth., 1997.

....Ve r e s t oy et al. 21] present a paper where several feature point trackers are used for particle image velocimetry. Among those trackers are the IPAN tracker [1] and the KanadeLucas Tomasi tracker [9, 18, 17] Point trackers have also been used to track facial features. McKenna et al. [11] proposed an approach to track rigid and non rigid facial motion based on a point distribution model (PDM) and Gabor wavelets. Petajan [12] uses facial feature tracking to track the eyes and the nostrils. Huang and Huang [2] also use a PDM approach to extract facial features. Their method measures ....

S. McKenna, S. Gong, R. P. Wurtz, J. Tanner, and D. Banin. Tracking facial feature points with Gabor wavelets and shape models. Proc. Int. Conf. on Audio- and VideoBased Biometric Person Authentication, Crans-Montana, Switzerland, pages 35--42, 1997.

.... be done by the means of active shape models using grey level models with gradient descend on the shape parameters [Luettin 97] or moving points along lines orthogonal to the curve [Cootes 92.2] Cootes 95] Cootes 96] active shape models using geometric histograms [Di Mauro 96] Gabor filters [McKenna 97], or the condensation algorithm using edges [Isard 96] In [Davis 97] and [DeCarlo 96] it is shown how to combine optical flow calculations with shape models (see [Barron 94] for an overview an evaluation of optical flow algorithms) However, optical flow algorithms are very sensitive to noise and ....

Stephen J. McKenna, Shaogang Gong, Rolf P. Wurtz, Jonathan Tanner, and Daniel Bain. "Tracking Facial Feature Points with Gabor Wavelets and Shape Models." In Proc. of the 1st Int. Conf. on Audio- and Videobased Biometric Person Authentication, Lecture Notes in Computer Science. 1997.

....the pyramidal system yields fairly dense and very accurate correspondence maps. The accuracy is mainly due to the phase adjustment. These can be used for tracking facial points. In this case, the coarse to fine matching becomes unpractical, and it has been replaced by a learned shape model in [4]. This sort of tracking can eventually lead to automatic measurement of emotions. 4. Recognition Once a mapping is established the feature similarities of corresponding point pairs are added up (or averaged) over the whole mapping. This yields a similarity value for each stored model, and the ....

Stephen J. McKenna, Shaogang Gong, Rolf P. Wurtz, Jonathan Tanner, and Daniel Banin. Tracking facial feature points with Gabor wavelets and shape models. In Proceedings of the First International Conference on Audio- and Video-based Biometric Person Authentication Crans-Montana, Switzerland, 12-14 March 1997, 1997. In press.

....Tracking feature points reliably through a sequence of images is a much desired skill for all applications where trajectories need to be measured and evaluated. In this context Gabor wavelets have turned out to be well suited to determine the disparity between two points from consecutive images [3, 4]. The phase of the complex response to a Gabor filter varies nearly linearly for small translations in the image plane [1] which allows disparity estimation with subpixel accuracy. Another important feature are the multi scale properties providing a very flexible point description and the ability ....

S. J. McKenna, S. Gong, R. P. Wurtz, J. Tanner, and D. Banin. Tracking facial feature points with Gabor wavelets and shape models. In J. Bigun, G. Chollet, and G. Borgefors, editors, Proc. 1st AVBPA, pages 35--42. Springer, 1997.

....a mean shape avoids the need to compute many correspondence maps. However, matching still entails a relatively expensive search over parameters of variations in shape and texture. The match can be based on models of the local appearance of the landmarks [38] or along curves joining the landmarks [14,22,25]. Crucially, linear active shape models are based on a number of implicit as sumptions: a) the shape of the object of interest can be defined by a relatively small set of explicit view models, b) the grey levels around a particular landmark are consistent for all the views of the object and ....

S. McKenna, S. Gong, R.P. Wiirtz, J. Tanner, and D. Banin. Tracking facial feature points with gabor wavelets and shape models. In IAPR Int. Conf. on Audio-Video Based Biometric Person Authentication, pages 35-43, CransMontana7 Switzerland7 1997.

....dense image correspondence is both problematic and expensive since changes in viewpoint could result in self occlusions which prohibit complete sets of image correspondences from being established. Alternatively, sparse correspondence can be established for a carefully chosen set of feature points [3]. A different approach that is computationally less expensive and does not rely on reliable and fast feature detection in every frame uses holistic appearance based templates. This assumes that all points of interest of an object move coherently in space and such an approximate rigidity assumption ....

....If image changes are approximately affine, correspondence for alignment can be achieved by treating an image 1 Throughout this paper, we use I to represent an image vector rather than the Identity matrix. Furthermore, a pre filtering process is often necessary if global illumination is unstable [3]. I(f(x; a) I(f(x 1 ; a) I(f(x 2 ; a) I(f(x N ; a) T as a function of an affine transform given by parameters a = a o ; a 1 ; a 2 ; a 3 ; a 4 ; a 5 ) T , where f (x; a) a 0 a 3 a 1 a 2 a 4 a 5 x Gamma x c y Gamma y c (1) x c = x c ; y c ) T is the ....

[Article contains additional citation context not shown here]

S. McKenna, S. Gong, R. Wurtz, J. Tanner, and D. Banin, "Tracking facial feature points with gabor wavelets and shape models," in AVBPA, Switzerland, 1997.

No context found.

S. J. McKenna, S. Gong, R. P. Wurtz, J. Tanner, D. Banin "Tracking Facial Feature Points with Gabor Wavelets and Shape Models" 1st Int. Conf. on Audio- and Videobased Biometric Person Authentication, Lecture Notes in Computer Science 1997.

No context found.

McKenna S,Gong S, Wu rtz R, Tanner J, Banin D.Tracking facial feature points with Gabor wavelets and shape models. In: International conference on audio and video-based biometric person authentication. Crans-Montana, Switzerland; 1997. p. 35--42.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC