D. DeCarlo, D. Metaxas, Optical flow constraints on deformable models with applications to face tracking, International Journal of Computer Vision 38 (2) (2000) 99--127.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....evaluation of different hypotheses. The way the two components are combined and weighted is application dependent and plays a decisive role in the robustness and efficiency of the tracker. For example, face tracking in a crowded scene relies more on target representation than on target dynamics [21], while in aerial video surveillance, e.g. 74] the target motion and the ego motion of the camera are the more important components. In real time applications, only a small percentage of the system resources can be allocated for tracking, the rest being required for the preprocessing stages or ....

D. DeCarlo and D. Metaxas, "Optical Flow Constraints on Deformable Models with Applications to Face Tracking," Int'l J. Computer Vision, vol. 38, no. 2, pp. 99-127, 2000.

....evaluation of different hypotheses. The way the two components are combined and weighted is application dependent and plays a decisive role in the robustness and efficiency of the tracker. For example, face tracking in a crowded scene relies more on target representation than on target dynamics [21], while in aerial video surveillance, e.g. 74] the target motion and the ego motion of the camera are the more important components. In real time applications only a small percentage of the system resources can be allocated for tracking, the rest being required for the preprocessing stages or ....

D. DeCarlo and D. Metaxas, "Optical flow constraints on deformable models with applications to face tracking," Intl. J. of Computer Vision, vol. 38, no. 2, pp. 99--127, 2000.

....range scanner. Some of them include information on how to move vertices according to physical properties of the skin and the underlying muscles [31] Another approach is to fit a generic 3 D model to image data by solving a dynamic system incorporating either optical flow and edge constraints [24] or stereo disparity maps [36] A technique for matching a 3 D model to a single photograph is proposed in [5] where the 3 D model is deformed in the directions of several principal components to fit the given two dimensional (2 D) image. To obtain a natural appearance, these 3 D models ....

....the base of the ears. Quads marking the mouth cheeks chin areas, eyes area, and forehead area are derived from these initial measurements. Since this is done only once, and only ten points are involved, there is little incentive to automate it. Techniques exist, such as the ones described in [17] [24], and [36] that can adapt a generic head model to a specific person from video sequences showing head movements. This may be useful if only video footage exists without the person being present. A person is then recorded while speaking freely in front of the camera. For the examples shown here, ....

D. DeCarlo and D. Metaxas, "Optical flow constraints, on deformable models with applications to human face shape and motion estimation," in Proc. CVPR, 1996, pp. 231--238.

....345 Scarborough Rd. Briarcli Manor, NY 10510. Email: mohamed.abdelmottaleb philips.com. Electrical and Computer Engineering Dept. Univ. of Miami. Email: mottaleb miami.edu. face recognition algorithms assume that the face location is known. Similarly, face tracking algorithms (e.g. [10]) often assume the initial face location is known. Note that face detection can be viewed as a two class (face vs. non face) classi cation problem. Therefore, some techniques developed for face recognition (e.g. holistic template approaches [12] 30] 41] 27] feature based approaches [6] ....

D. DeCarlo and D. Metaxas, \Optical Flow Constraints on Deformable Models with Applications to Face Tracking," Int'l Journal Computer Vision, vol. 38, no. 2, pp. 99-127, July 2000.

.... expression components [Parke82] Duffy88] Thalm89] Pighn98] For physics based face models, animation can be performed using numerical simulation of muscle actions [Water87] Lee95] Many face tracking algorithms have been devised including those based on optical flow constraints [Mase91][DeCar00], active shape models (ASM) Baumb96] Edwar98] or energy minimising point tracking techniques [Lucas81] Lien00] The reconstruction of tracked facial feature movements by virtual actors has application in video telecommunication because of the potential for low bandwidth communication [Choi91] ....

DeCarlos D. and Metaxas D.: Optical flow constraints on deformable models with applications to face tracking, International Journal of Computer Vision, Vol. 38, No. 2, pp99-127, 2000.

....of applications that require a multimodal interface with the virtual environment has steadily increased. Within this field of research, recognition of the facial expressions is a very complex and interesting subject where there have been numerous research efforts. For instance, DeCarlo et al. [1] have applied optical flow and a generic face model based algorithm. This method is robust but it takes much time to recognize the face, and it is not in real time. Cosatto et al. 2] used a sample based method which needs to make a sample for each person. Kouadi et al. 3] used a database and ....

D. DeCarlo, D. Metaxas, `Optical Flow Constraints on Deformable Models with Applications to Face Tracking', CIS technical report MS-CIS-97-23

....application that requires a fully multimodal interface with the virtual environment has steadily increased. Within this field of research, recognition of facial expressions is a very complex and interesting subject where there have been numerous research efforts. For instance, DeCarlo and Metaxas [1] have applied optical flow and a generic face model based algorithm. This method is robust but it takes much time to recognize face, and it is not in real time. Cosatto and Graf [2] used a sample based method. This method needs to make a sample for each person. Kouadi et al. 3] also used sample ....

....system will be described in details. The paper will conclude on real time results applied to a compatible facial animation system. 2. System Overview Figure 1 sketches the different tasks and interactions to generate a real time virtual dialog between a synthetic clone and an autonomous actor [1]. The video and the speech of the user drive the clone facial animation, while the autonomous actor uses the information of speech and facial emotions from the user to generate an automatic behavioral response. MPEG 4 Facial Animation Parameters are extracted in real time from the video input of ....

D. DeCarlo, D. Metaxas, `Optical Flow Constraints on Deformable Models with Applications to Face Tracking', CIS technical report MS-CIS-97-23

No context found.

D. DeCarlo and D. Metaxas. Optical flow constraints on deformable models with applications to face tracking. International Journal of Computer Vision, 38(2):99--127, 2000.

....system. The next description describes this implemented system, as well as its specific model for partial computations. 5 Implementation The domain of tracking a human face is used as a testbed for our system. This is constructed in part from a previously developed face tracking system [30], the relevant details of which follow. 5.1 Face Model A three dimensional face model is used to describe the shape, motion and appearance of faces in images. The model is formed by applying deformation functions to an underlying shape (a polygon mesh representing an average face) Within our ....

....the face model undergoing various shape deformations (showing four di#erent individuals) motion deformations (showing brow raising and frowning, smiling, and mouth opening) and finally two examples of when several deformations are applied at once. Further detail about this model can be found in [30]. The next section describes the particular cue computations used by this system, to estimate shape and motion. 5.2 Cue Computations Model based optical flow. This computation is an iterative algorithm which solves a non linear least squares problem to determine the motion estimate of the face ....

[Article contains additional citation context not shown here]

DeCarlo, D., Metaxas, D.: Optical flow constraints on deformable models with applications to face tracking. IJCV 32 (2000) 99--127

....of the shading of the object as it rotates with respect to the light source. This constraint unifies the shading and the optical flow constraints and degenerates to each one of them when the other is not present. Although optical flow and edges in deformable models have been used in the past [18], as well as shading [17] these two methods were applied to different problem domains (moving and static objects respectively) In this paper we combine them to correct for the errors due to the brightness constancy assumption. We use cue information from the entirety of the hand and we are able ....

....q) 0 (6) We notice that if J is the Jacobian of the model points, and J p is the Jacobian of the model points under perspective projection, as described in Sec. 3, then = #ILJ p J q (7) is the left hand side of the model based optical flow constraint equation that was presented in [18]. As explained in this paper, in model based optical flow, motion field vectors are vectors of velocities of model points, and hence x = J q applies. Typically in the literature [11] this optical flow term is set to 0. This is correct in the case of ambient only illumination. For the case of ....

D. DeCarlo and D. Metaxas. Optical Flow Constraints on Deformable Models with Applications to Face Tracking. IJCV, July 2000, 38:2, pp. 99-127.

....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 secondary cue. This framework requires an a priori user based definition of which cue is the most ....

....force in parameter space and to average them into a single generalized force f g , which is then used in Equation 1. This approach, however, completely ignores that the estimates from the cues exhibit different degrees of reliability. An alternative approach is to use one cue as a constraint [14], but this approach is still too inflexible in the case when the constraining cue becomes unreliable. A statistical approach seems like the obvious solution, but it opens another can of worms: how do we estimate the probability distributions of the contributions from the various image forces ....

[Article contains additional citation context not shown here]

D. DeCarlo and D. Metaxas. Optical flow constraints on deformable models with applications to face tracking. Int. J. of Comp. Vision, 38(2):99--127, July 2000.

....sources. We confirm this assumption empirically in the context of face 4 tracking in Section 5.2, and it is likely to apply in other domains where tracking with model based optical flow is successful. The work in this paper builds on the model based face tracking framework described in [7]. This framework uses a model based optical flow computation as a constraint on the motion of a deformable model, which uses features (edges) to align the model with the image. Using features prevents the accumulation of tracking error, which would have otherwise been a difficulty using flow ....

....point on the model is written as x(q;u) with u # W, although the dependency of x on q is often omitted. The goal of shape and motion estimation is to recover the value of q over time from a sequence of images. For this paper, we will be using the three dimensional parameterized face model from [7]. 5 As stated earlier, to distinguish the processes of shape estimation and motion tracking, the parameters in q are rearranged and separated into q b (the basic shape of the object) and qm (rigid and non rigid motion) so that q = q # b , q # m ) # . Within our face model, q b describes an ....

[Article contains additional citation context not shown here]

D. DeCarlo and D. Metaxas. Optical flow constraints on deformable models with applications to face tracking. International Journal of Computer Vision, 32(2):99--127, July 2000.

....a probability distribution. As a result, the optimal integration of cues to yield the best possible parameter estimate of the model is a difficult and open research problem. Previous approaches integrated the cues either by using a direct sum of the cues, or through the design of hard constraints [4] that subjugate some cues to others. A direct sum ignores that some cues may be more reliable than others at a given point in time, whereas a hard constraint causes problems if the dominant cue is unreliable or changes over time. In this paper we describe a new method for combining the cues ....

....in experiments on synthetic images and real face tracking data that the statistical approach is more robust than the simple direct sum of the cues. 1.1. Related Work Deformable Models have been used in a variety of areas and applications. In computer vision for tracking and shape estimation [8, 4, 24], in computer graphics for synthesis and simulation [6] and in medical applications for reconstruction, modeling and diagnosis [22, 2] Most of these approaches have been deterministic; that is, they did not address the statistical uncertainties inherent in tracking images, and in fitting the ....

D. de Carlo and D. Metaxas. Optical flow constraints on deformable models with applications to face tracking. IJCV, 38(2):99--127, July 2000.

No context found.

D. DeCarlo, D. Metaxas, Optical flow constraints on deformable models with applications to face tracking, International Journal of Computer Vision 38 (2) (2000) 99--127.

No context found.

D. DeCarlo and D. Metaxas. Optical flow constraints on deformable models with applications to face tracking. IJCV, 38(2):99--127, 2000.

No context found.

D. DeCarlo and D. Metaxas. Optical flow constraints on deformable models with applications to face tracking. Int'l J. of Computer Vision, 38(2):99--127, July 200.

No context found.

D. de Carlo and D. Metaxas. Optical flow constraints on deformable models with applications to face tracking. IJCV, 38(2):99--127, July 2000.

No context found.

D. DeCarlo and D. Metaxas. Optical flow constraints on deformable models with applications to face tracking. IJCV, 38(2):99--127, 2000.

No context found.

Douglas Decarlo and Dimitri Metaxas. Optical flow constraints on deformable models with applications to face tracking. International Journal of Computer Vision (IJCV), 38(2):99--127, 2000.

No context found.

Douglas DeCarlo and Dimitris Metaxas. Optical flow constraints on deformable models with applications to face tracking. International Journal of Computer Vision, 38(2):99--127, 2000.

No context found.

Douglas DeCarlo and Demetri Metaxas. Optical flow constraints on deformable models with applications to face tracking. International Journal of Computer Vision, 38:99--127, February 2000.

No context found.

D. DeCarlo and D. Metaxas, "Optical flow constraints on deformable models with applications to face tracking," IJCV, vol. 38, no. 2, pp. 99--127, July 2000.

No context found.

D. DeCarlo and D. Metaxas, "Optical flow constraints on deformable models with applications to face tracking," International Journal of Computer Vision, vol. 38, no. 2, pp. 99--127, July 2000.

No context found.

D. DeCarlo and D. Metaxas. Optical flow constraints on deformable models with applications to face tracking. In International Journal of Computer Vision. 2000. 38(2):99-127.

No context found.

D. DeCarlo and D. N. Metaxas. Optical flow constraints on deformable models with applications to face tracking. IJCV, 38(2):99--127, 2000.

No context found.

D. DeCarlo and D. N. Metaxas. Optical flow constraints on deformable models with applications to face tracking. IJCV, 38(2):99--127, 2000.

No context found.

D. DeCarlo and D. Metaxas. Optical flow constraints on deformable models with applications to face tracking. IJCV, 38(2):99--127, 2000.

No context found.

D. DeCarlo and D. Metaxas. Optical flow constraints on deformable models with applications to face tracking. International Journal in Computer Vision, 38(2):99-- 127, July 2000.

No context found.

Douglas DeCarlo and Dimitris Metaxas. Optical Flow Constraints on Deformable Models with Applications to Face Tracking. International Journal of Computer Vision, 38(2):99--127, July 2001.

No context found.

D. DeCarlo and D. Metaxas, "Optical flow constraints on deformable models with applications to face tracking," Intl. J. of Computer Vision, vol. 38, no. 2, pp. 99--127, 2000.

No context found.

Douglas DeCarlo and Demetri Metaxas. Optical flow constraints on deformable models with applications to face tracking. International Journal of Computer Vision, 38:99--127, February 2000.

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