F. Pighin, R. Szeliski, and D. H. Salesin. Resynthesizing facial animation through 3D model-based tracking. In Proc., ICCV, pages 143--150, Corfu, Greece, 1999. IEEE Computer Society.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....from shading obtain initial fits [36] and using a similar framework, 15] uses anthropometric data and inspired deformations to generate faces. A learning based statistical model can help tracking of face models [6] Eigen based approaches can successfully track, fit, and even recognize objects [33, 5, 34]. In [9] the head is modeled as a cylinder, and in [3] as a plane, and in [34] tracking is used for animation, to mention just a few. Cue integration is not a new topic. In [14] a two cue integration algorithm is presented based on the use of constraints, in which optical flow is defined to be ....

....data and inspired deformations to generate faces. A learning based statistical model can help tracking of face models [6] Eigen based approaches can successfully track, fit, and even recognize objects [33, 5, 34] In [9] the head is modeled as a cylinder, and in [3] as a plane, and in [34] tracking is used for animation, to mention just a few. Cue integration is not a new topic. In [14] a two cue integration algorithm is presented based on the use of constraints, in which optical flow is defined to be the constraining (i.e. most important) cue, and edges are defined to be the ....

[Article contains additional citation context not shown here]

F. Pighin, R. Szeliski, and D. Salesin. Resynthesizing facial animation through 3d model-based tracking. In ICCV, pages 143--150, 1999.

....we have created a simple, yet powerful, deformable face model using the dag data structure. We show some results of our face tracking system. We started with a static a geometric model of a head, publicly made available by the computer graphics group of the University of Washington as part of [16, 17]. We carefully extracted a mask of the face, and simpli ed it using SlimKit Surface Modeling Tools . The result was a static mask model of a generic face with 1101 nodes and 2000 faces. We then marked regions in the surface that are a ected by a parameter, and how this e ect varied in this ....

F. H. Pighin, R. Szeliski, and D. Salesin. Resynthesizing facial animation through 3d model-based tracking. In ICCV, pages 143150, 1999.

....how to exploit it for the two step tracking. In section 5 we summarize our experiments and we conclude by discussing the results in section 6. 2 Previous Work Many non rigid tracking solutions have been proposed previously. As mentioned earlier, most methods use an a priori model. Examples are [11, 3, 5, 14, 1, 2, 10]. Most of these approaches estimate non rigid 2D motion, but some of them also recover 3D pose and deformations based on a 3D model. What is most closely related to our approach, and in part inspired this solution, is work by Irani and Anandan [8, 9] as well as methods for non rigid ....

F. Pighin, D. H. Salesin, and R. Szeliski. Resynthesizing facial animation through 3d model-based tracking. In ICCV, 1999.

....unified manner. We demostrate the technique on tracking several video sequences and on deriving 3D deformable models from those measurements. 2 Previous Work Many non rigid tracking solutions have been proposed previously. As mentioned earlier, most techniques use an a priori model. Examples are [16, 5, 9, 19, 3, 4]. Most of these techniques model 2D non rigid motion, but some of these approaches also recover 3D pose and deformations based on a 3D model. The 3D model is obtained from 3D scanning devices [6] stereo cameras [10] or multi view reconstruction [18, 11] The multi view reconstruction is based on ....

F. Pighin, D. H. Salesin, and R. Szeliski. Resynthesizing facial animation through 3d model-based tracking. In ICCV, 1999.

....3D model of the shape space was obtained by laser scanning a large face database a priori. Using a hand initialization and iterative matching of shape, texture, and lighting, a very detailed 3D face shape could be recovered from one single image. Based on 2D image sequences, 6] and [10] were tracking the pose and configuration of human faces. A 3D face model was given a priori as well. Basu [2] demonstrates how the parameters can be iteratively fitted to a video sequence, starting from an initial lip model. 11, 7] propose methods for recovering the 3D facial model itself using ....

F. Pighin, D. H. Salesin, and R. Szeliski. Resynthesizing facial animation through 3d model-based tracking. In Proc. Int. Conf. Computer Vision, 1999.

....model of face shape and texture can be used to generate new views given a single view. The model can be matched to a new image from more or less any viewpoint using a general optimisation scheme, though this is slow. Similar work has been described by Fua and Miccio [26] and Pighin et al.[56]. By explicitly taking into account the 3D nature of the problem, this approach is likely to yield better reconstructions than the purely 2D method described below. However, the view based models we propose could be used to drive the parameters of the 3D head model, speeding up matching times. ....

F. Pighin, R. Szeliski, and D. Salesin. Resynthesizing facial animation through 3d model-based tracking. In 7 th International Conference on Computer Vision, pages 137-142, 1999.

....can represent this variability. For instance, the majority of work on face tracking and recognition assumes near fronto parallel views, and tends to break down when presented with large rotations or pro le views. Three general approaches have been used to deal with this; a) use a full 3D model [64, 18, 51], b) introduce non linearities into a 2D model [29, 53, 60] and c) use a set of models to represent appearance from di erent view points [45, 35, 66] In this chapter we explore the last approach, using statistical models of shape and appearance to represent the variations in appearance from a ....

....model of face shape and texture can be used to generate new views given a single view. The model can be matched to a new image from more or less any viewpoint using a general optimization scheme, though this is slow. Similar work has been described by Fua and Miccio [18] and Pighin et al.[51]. By explicitly taking into account the 3D nature of the problem, this approach is likely to yield better reconstructions than the purely 2D method described below. However, the view based models we propose could be used to drive the parameters of the 3D head model, speeding up matching times. La ....

F. Pighin, R. Szeliski, and D. Salesin. Resynthesizing facial animation through 3d model-based tracking. In 7 th International Conference on Computer Vision, pages 143-150, 1999.

....of a face as seen from two di erent view points. The majority of work on face tracking and recognition assumes near fronto parallel views, and tends to break down when presented with large rotations or pro le views. Three general approaches have been used to deal with this; a) use a full 3D model [18, 4, 13], b) introduce non linearities into a 2D model [6, 14, 16] and c) use a set of models to represent appearance from di erent view points [12, 2] In this paper we explore the last approach, using statistical models of shape and appearance to represent the variations in appearance from a particular ....

....model of face shape and texture can be used to generate new views given a single view. The model can be matched to a new image from more or less any viewpoint using a general optimisation scheme, though this is slow. Similar work has been described by Fua and Miccio [4] and Pighin et.al. [13]. By explicitly taking into account the 3D nature of the problem, this approach is likely to yield better reconstructions than the purely 2D method described below. However, the view based models we propose could be used to drive the parameters of the 3D head model, speeding up matching times. 3 ....

F. Pighin, R. Szeliski, and D. Salesin. Resynthesizing facial animation through 3d model-based tracking. In 7 th International Conference on Computer Vision, pages 137-142, 1999.

No context found.

F. Pighin, R. Szeliski, and D.H. Salesin. Resynthesizing facial animation through 3d model-based tracking. In Proceedings, International Conference on Computer Vision, 1999.

....in this dissertation is largely drawn from two publications. The first one was published in the Proceedings of the ACM SIGGRAPH Conference in 1998 [60] and covers Chapters 2, 3, and 4. The second one was published in the Proceedings of the International Conference on Computer Vision in 1999 [61] and covers Chapter 5 . 16 Chapter 2 MODELING FROM PHOTOGRAPHS 2.1 Introduction Creating realistic face models is a very challenging problem. The human face is an extremely complex geometric form. Moreover, the face exhibits countless tiny creases and wrinkles, as well as subtle variations in ....

F. Pighin, R. Szeliski, and D.H. Salesin. Resynthesizing facial animation through 3d model-based tracking. In Proceedings, International Conference on Computer Vision,1999.

....42] systems requiring special markers or devices are unlikely to be adopted by the casual user. Using vision based techniques, facial features can be tracked non invasively. Trackers can be based on deformable patches [8, 14] edge or feature detectors [9, 26, 24, 43, 34, 53] and or 3D models [3, 17, 23, 45]. Face tracking is currently an active area of research: robust, full featured, real time face tracking remains elusive. In this paper, we use a simple, color based feature tracker (Section 2) that runs in real time. On the rendering side, 3D facial animation is one of the most actively studied ....

Frederic Pighin, Richard Szeliski, and David H. Salesin. Resynthesizing facial animation through 3D model-based tracking. In Seventh IEEE International Conference on Computer Vision (ICCV '99), pages 143--150, 1999.

No context found.

F. Pighin, R. Szeliski, and D. H. Salesin. Resynthesizing facial animation through 3D model-based tracking. In Proc., ICCV, pages 143--150, Corfu, Greece, 1999. IEEE Computer Society.

No context found.

F. Pighin, R. Szeliski, and D. Salesin. Resynthesizing facial animation through 3d model-based tracking. In International Conference on Computer Vision. 1999. 143-150.

No context found.

F. Pighin, R. Szeliski, and D. Salesin. Resynthesizing facial animation through 3d model-based tracking. In 7 International Conference on Computer Vision, pages 143--150, 1999.

No context found.

F. Pighin, R. Szeliski, and D. Salesin. Resynthesizing facial animation through 3d modelbased tracking. In 7 on Computer Vision, pages 143-150, 1999.

No context found.

F. Pighin, R. Szeliski, and D. H. Salesin. Resynthesizing facial animation through 3D model-based tracking. In Seventh IEEE International Conference on Computer Vision (ICCV '99), pages 143--150, 1999.

No context found.

F. Pighin, R. Szeliski, and D. Salesin. Resynthesizing facial animation through 3d model-based tracking. In 7 ############# ########## ## ######## ######, pages 137-142, 1999.

No context found.

F. Pighin, D. H. Salesin, and R. Szeliski. Resynthesizing facial animation through 3d model-based tracking. In ICCV, 1999.

No context found.

F. Pighin, R. Szeliski, and D. H. Salesin. Resynthesizing facial animation through 3D model-based tracking. In Proc., ICCV, pages 143--150, Corfu, Greece, 1999. IEEE Computer Society.

No context found.

F. Pighin, R. Szeliski, and D. H. Salesin. Resynthesizing facial animation through 3D model-based tracking. In Seventh International Conference on Computer Vision (ICCV '99) Conference Proceedings, pages 143--150, September 1999. Corfu, Greece.

No context found.

F. Pighin, D. H. Salesin, and R. Szeliski, Resynthesizing facial animation through 3d model-based tracking., in In Seventh International Conference on Computer Vision (ICCV'99), pp. 143150, (Kerkyra, Greece), 1999.

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