T. Akimoto, Y. Suenaga, and R. Wallace. Automatic creation of 3d facial models. IEEE Computer Graphics and Applications, 13(5):16---22, 1993.  Home/Search   Document Not in Database   Summary   Related Articles   Check

.... virtual view. Figure from [11] knowledge of a facial transformation such as head rotation or expression change. A standard approach used in the computer graphics and computer vision communities for representing this prior knowledge is to use a 3D model of the face (Akimoto, Suennaga, and Wallace[3], Waters and Terzopoulos[36] 39] Aizawa, Harashima, and Saito[1] Essa and Pentland [20] After the single available 2D image is texture mapped onto a 3D polygonal or multilayer mesh model of the face, rotated views can be synthesized by rotating the 3D model and rendering. In addition, facial ....

Takaaki Akimoto, Yasuhito Suenaga, and Richard S. Wallace. Automatic creation of 3D facial models. IEEE Computer Graphics and Applications, 13(5):16--22, 1993.

....DiPaola s Facial Animation System (FAS) 28] is an extension of Parke s approach. New facial models are generated by digitizing live subjects or sculptures, or by manipulating existing models with free form deformations, stochastic noise deformations, or vertex editing. Akimoto, et al. [3] uses front and profile images of a subject to automatically create 3D facial models. Additional fitting techniques are described in [58, 118, 137] 11. Animation using Tracking Fig. 11 Real time tracking is performed without markups on the face using Eyematic Inc. s face tracking system. ....

T. Akimoto, Y. Suenaga, R. Wallace, Automatic creation of 3D facial models. IEEE computer Graphics and Application, 1993, vol. 13(5), pp. 16-22

....It is sufficient to use rough estimates of their values as a starting point of the automated matching procedure. Most techniques for face cloning , the reconstruction of a 3D face model from one or more images, still rely on manual assistance for matching a deformable 3D face model to the images [26, 1, 30]. The approach of Pighin et al. 28] demonstrates the high realism that can be achieved for the synthesis of faces and facial expressions from photographs where several images of a face are matched to a single 3D face model. Our automated matching procedure could be used to replace the manual ....

T. Akimoto, Y. Suenaga, and R.S. Wallace. Automatic creation of 3D facial models. IEEE Computer Graphics and Applications, 13(3):16--22, 1993.

....these methods correspond to the VFP visibility problems that will be discussed in the next chapter. For the purpose of ray tracing acceleration, methods that aim to reduce the number of rays to be traced have been proposed. Firstly, the coherence of radiance along similar rays can be utilised [3, 111, 104, 186]. Secondly a visibility predetermination can avoid casting rays between invisible regions (points) 85, 207, 114] Further acceleration of ray shooting can be achieved by speeding up the ray object intersection itself. This is usually carried out by object bounding volumes such as simple axis ....

T. Akimoto, K. Mase, and Y. Suenaga. Pixel-selected ray tracing. IEEE Computer Graphics and Applications, 11(4):14--22, July 1991.

....camps, both crucially dependent (in different ways) on human participation. The first approach is to extract geometry automatically from the measurement of a live subject. Lee, et al. LTK95] use a range scan of a subject, and produce a physics based model capable of animation. Akimoto, et al. ASR93] use front and profile images of a subject to produce a model. The second approach is to facilitate manual specification of new face geometry by a user. A certain facility is offered already by commercial modelers (though of course their use demands considerable artistic skill) several ....

T. Akimoto, Y. Suenaga, and R.Wallace. Automatic creation of 3D facial models. IEEE Computer Graphics and Applications, 13(5):16--22, September 1993.

....projection in the new view. Previous authors have taken only a subset of these considerations into account when designing their weighting functions. For example, Kurihara and Arai [43] use positional certainty as their weighting function, but they do not account for selfocclusion. Akimoto et al. [1] and Ip and Yin [38] blend the images smoothly, but address neither self occlusion nor positional certainty. Debevec et al. 17] who describe a view dependent texture mapping technique for modeling and rendering buildings from photographs, do address occlusion but do not account for positional ....

T. Akimoto, Y. Suenaga, and R.S. Wallace. Automatic creation of 3D facial models. IEEE Computer Graphics and Applications, 13(5):16--22, September 1993.

....face models based on face measurements randomly generated according to anthropometric statistics. They showed that they were able to generate a variety of face geometries using these face measurements as constraints. A number of researchers have proposed to create face models from two views [1, 16, 4]. They all require two cameras which must be carefully set up so that their directions are orthogonal. Zheng [40] developed a system to construct geometrical object models from image contours, but it requires a turn table setup. Pighin et al. 29] developed a system to allow a user to manually ....

T. Akimoto, Y. Suenaga, and R. S. Wallace. Automatic 3d facial models. IEEE Computer Graphics and Applications, 13(5):16--22, September 1993.

....in the new view. Previous authors have taken only a subset of these considerations into account when designing their weighting functions. For example, Kurihara and Arai [24] use positional certainty as their weighting function, but they do not account for self occlusion. Akimoto et al. [1] and Ip and Yin [20] blend the images smoothly, but address neither self occlusion nor positional certainty. Debevec et al. 9] who describe a view dependent texture mapping technique for modeling and rendering buildings from photographs, do address occlusion but do not account for positional ....

....number of starting expressions by taking convex combinations of them all, using weights that apply both to the coordinates of the mesh vertices and to the values in the texture map. Extrapolation of expressions should also be possible by allowing weights to have values outside of the interval [0, 1]; note, however, that such weights might result in colors outside of the allowable gamut. We can generate an even wider range of expressions using a localized blend of the facial expressions. Such a blend is specified by a set of blend functions, one for each expression, defined over the ....

Takaaki Akimoto, Yasuhito Suenaga, and Richard S. Wallace. Automatic Creation of 3D Facial Models. IEEE Computer Graphics and Applications, 13(5):16--22, September 1993.

.... virtual view. Figure from [11] knowledge of a facial transformation such as head rotation or expression change. A standard approach used in the computer graphics and computer vision communities for representing this prior knowledge is to use a 3D model of the face (Akimoto, Suennaga, and Wallace[3], Waters and Terzopoulos[36] 39] Aizawa, Harashima, and Saito[1] Essa and Pentland [20] After the single available 2D image is texture mapped onto a 3D polygonal or multilayer mesh model of the face, rotated views can be synthesized by rotating the 3D model and rendering. In addition, facial ....

Takaaki Akimoto, Yasuhito Suenaga, and Richard S. Wallace. Automatic creation of 3D facial models. IEEE Computer Graphics and Applications, 13(5):16--22, 1993.

....poseexpression lighting parameters. For synthesizing images of faces, 3D facial models have been explored in the computer graphics, computer vision, and model based image coding communities (Aitchison and Craw[1] Kang, Chen, and Hsu[24] Essa and Pentland [19] Akimoto, Suennaga, and Wallace[3], Waters and Terzopoulos[47] Aizawa, Harashima, and Saito[2] In the 3D technique, face shape is represented either by a polygonal model or by a more complicated multilayer mesh that simulates tissue. Once a 2D face image is texture mapped onto the 3D model, the face can be treated as a ....

Takaaki Akimoto, Yasuhito Suenaga, and Richard S. Wallace. Automatic creation of 3D facial models. IEEE Computer Graphics and Applications, 13(5):16--22, 1993.

....angles. A common technique is to place markers on the face that can be seen from two or more cameras. An alternative technique is to manually digitize a plaster cast of the face using manual 3D digitization devices such as orthogonal magnetic fields sound captors [9] or one to two photographs [9, 7, 1]. More recently, automated laser range finders can digitize on the order of 10 5 3D points from a solid object such as a person s head and shoulders in just a few seconds [23] In procedure (3) an animator must decide which mesh nodes to articulate and how much they should be displaced in order ....

T. Akimoto, Y. Suenaga, and R. Wallace. Automatic creation of 3D facial models. IEEE Computer Graphics and Applications, 13(5):16-- 22, September 1993.

....face [25, 26] One consequence of using these grids, however, is that the images used to construct geometry can no longer be used as a valid texture maps for the subject. More recently, several methods have been proposed for modeling the face photogrammetrically without the use of grids [19, 1, 16]. These modeling methods are quite similar in their basic concept to the modeling technique described in this paper. However, we improve upon these previous methods in a number of ways, including allowing fairly arbitrary camera positions and lenses (rather than using a fixed pair that are ....

....image and the direction of projection in the new view. Previous authors used weighting function that satisfy only a subset of these requirements. For example, Kurihara and Arai [19] use positional uncertainty as their weighting function, but they do not account for selfocclusion. Akimoto et al. [1] as well as Ip and Yin [16] blend the images smoothly, but address neither self occlusion nor positional certainty. Debevec et al. 7] who describe a view dependent texture mapping technique for modeling and rendering buildings from photographs, do address occlusion but do not account for ....

Takaaki Akimoto, Yasuhito Suenaga, and Richard S. Wallace. Automatic creation of 3D facial models. IEEE Computer Graphics and Applications, 13(5):16--22, September 1993.

No context found.

T. Akimoto, Y. Suenaga, and R. Wallace. Automatic creation of 3d facial models. IEEE Computer Graphics and Applications, 13(5):16---22, 1993.

....by two criteria: area coverage and image feature location. If a region reaches the size of a pixel, the only criterion used is mean variance. The refinement process stops when a particular confidence level of the image intensity is reached, even though not all pixels have been sampled. Other [1, 6, 14] sample generators have been proposed for producing an optimal sampling pattern. The sample location generator of Eldar et al. 6] designed for image compression) maintains a growing Delaunay triangulation [17] of sample locations. These triangles are continuously refined. A new sample location ....

T. Akimoto, K. Mase, and Y. Suenaga. Pixel-selected ray tracing. IEEE Computer Graphics and Applications, 11(4):14--22, 1991.

No context found.

T. Akimoto, Y. Suenaga, and R. Wallace. Automatic creation of 3d facial models. IEEE Computer Graphics and Applications, 13(5):16---22, 1993.

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