BREGLER, C., AND OMOHUNDRO, S. 1995. Nonlinear image interpolation using manifold learning. In Advances in Neural Information Processing Systems 7. 973--980.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....width and height, etc. to the controls of a polygonal lip model [1] Still others have a trained model of lip variations and attempt to fit the observations to this model. Some of the most interesting work done in this area has been along these lines: Bregler and Omohundro s work, for example [5], models the non7 linear subspace of valid lip poses within the image space and can thus be used for both analysis and synthesis. Similarly, Luettin s system learns the subspace of variations for 2D contours surrounding the lips [11] However, in order for these 2D models to be robust, they have ....

Christoph Bregler and Stephen M. Omohundro. "Nonlinear Image Interpolation using Manifold Learning". In Advances in Neural Information Processing Systems 7,1995.

....calculated for an unknown object at an unknown pose; the test input is thus represented by a point in feature space. Its class is determined by the FST that is closest to it, and its pose is given by the closest line segment on that FST (Figure 5. 1) The FST concept was developed independently in [78], 79] and [27] In [79] the emphasis was on classification (not pose estimation) and on the use of a complex CHAPTER 5. MRDFS FOR CLASSIFICATION AND POSE ESTIMATION 140 learning manifold between training samples. In [27] the emphasis was active vision, where the sensor is moved to the best ....

C. Bregler and S. Omohundro, "Nonlinear image interpolation using manifold learning," in Proceedings of NIPS94 - Neural Information Processing Systems: Natural and Synthetic, vol. 3073, pp. 973--980, Nov. 1994.

....models are only two dimensional. Many are based directly on image data [5] others use such low level features to form a parametrized description of the lip shape [1] Some of the most interesting work done in this area has been in using a statistically trained model of lip variations (such as [4]) However, since these are 2D models, the changes in the apparent lip shape due to rigid rotations have to be modeled as complex changes in the lip pose. In our work, we begin by extending this philosophy to 3D. The other category of lip models includes those designed for synthesis and ....

Christoph Bregler and Stephen M. Omohundro. "Nonlinear Image Interpolation using Manifold Learning". In G. Tesauro, D. Touretzky, and T. Leen, editors, Advances in Neural Info. Proc. Sys. 7, pages 401--408, Cambridge, MA, 1995. MIT Press.

....faces) tends to be highly nonlinear. Fortunately, despite their complicated global structure, we can usually characterize these manifolds as locally linear. Thus, to a good approximation, they can be represented by collections of simpler models, each of which describes a locally linear neighborhood[3, 6, 8]. For unsupervised learning tasks, a probabilistic model that nicely captures this intuition is a mixture of factor analyzers (MFA) 5] The model is used to describe high dimensional data that lies on or near a lower dimensional manifold. MFAs parameterize a joint distribution over observed and ....

C. Bregler & S. Omohundro. Nonlinear image interpolation using manifold learning.

....to minimize the strain felt throughout the linear elastic FEM structure. It is worth noting that the tracking system uses normalized or chromatic color information (r = r (r g b) to make it robust against variations in lighting conditions. The goal was to extend similar ideas from 2D [15, 67] to 3D structure. The 2D models in [15, 67] suffered from 13 Rapid sequences of speech require that the posture for one phoneme anticipate the posture for the next phonemes. Conversely, the posture for the current phoneme is modified by the previous phonemes. This overlap between phonetic ....

....the linear elastic FEM structure. It is worth noting that the tracking system uses normalized or chromatic color information (r = r (r g b) to make it robust against variations in lighting conditions. The goal was to extend similar ideas from 2D [15, 67] to 3D structure. The 2D models in [15, 67] suffered from 13 Rapid sequences of speech require that the posture for one phoneme anticipate the posture for the next phonemes. Conversely, the posture for the current phoneme is modified by the previous phonemes. This overlap between phonetic segments is referred to as co articulation [55] ....

C. Bregler, Stephen, M. Omohundro, Nonlinear Image Interpolation using Manifold Learning, In NIPS 7, 1995

....A complete discussion is not possible, so the sequel rather focuses on a number of contributions that are particularly relevant for the method presented here. So far, for reaching photorealism one of the most effective approaches has been the use of 2D morphing between photographic images [9, 2, 6]. These techniques typically require animators to specify carefully chosen feature correspondences between frames. Bregler et al. 5] used morphing of mouth regions to lip synch existing video to a novel sound track. This Video Rewrite approach works largely automatically and directly from speech. ....

C. Bregler and S. Omohundro. Nonlinear image interpolation using manifold learning. In NIPS, volume 7, 1995.

....faces) tends to be highly nonlinear. Fortunately, despite their complicated global structure, we can usually characterize these manifolds as locally linear. Thus, to a good approximation, they can be represented by collections of simpler models, each of which describes a locally linear neighborhood[3, 6, 8]. For problems in unsupervised learning, a probabilistic model that nicely captures this intuition is a mixture of factor analyzers (MFA) 5] The model is used to describe high dimensional data that lies on or near a lower dimensional manifold. MFAs parameterize a joint distribution over observed ....

C. Bregler & S. Omohundro. Nonlinear image interpolation using manifold learning. Advances in Neural Information Processing Systems 7 (1995).

.... including view morphing [12] plenoptic modeling depth recovery [8] lightfields [7] and recent approaches using the trifocal tensor for view extrapolation [13] For non rigid view synthesis, networks for model based interpolation and manifold learning have been used successfully in some cases [14, 2, 4, 11]. Techniques based on Radial Basis Function (RBF) interpolation or on Principle Components Analysis (PCA) have been able to interpolate face images under varying pose, expression and identity [1, 5, 6] However, these methods are limited in the types of object appearance they can accurately ....

....we find local regions of appearance which are well behaved as smooth, possibly linear, functions. We wish to cluster our examples into sets which can be used for successful interpolation using our local appearance model. Conceptually, this problem is similar to that faced by Bregler and Omohundro [2], who built image manifolds using a mixture of local PCA models. Their work was limited to modeling shape (lip outlines) they used K means clustering of image appearance to form the initial groupings for PCA analysis. However this approach had no model of texture and performed clustering using a ....

C. Bregler and S. Omohundro, Nonlinear Image Interpolation using Manifold Learning, NIPS-7, MIT Press, 1995.

....map (SOM; Kohonen, 1988) the generative topographic mapping (GTM; Bishon, Svensen, Williams, 1998) or autoencoder neural networks (DeMers Cottrell, 1993) try to generalize PCA by discovering a single global low dimensional nonlinear model of the observations. In contrast, local methods (Bregler Omohundro, 1995; Hinton, Revow, Dayan, 1995) seek a set of low dimensional models, usually linear and hence valid only for a limited range of data. When appropriate, a single global model is 1 Given by x 1 = z 1 cos(z 1 ) x 2 = z 1 sin(z 1 ) x 3 = z 2 , for z 1 2 [3=2; 9=2] z 2 2 [0; 15] 0 10 0 10 0 ....

....can map the transformation between two images x (1) and x (2) onto an analogous transformation of another image x (3) by adding the transformation vector (y (2) Gamma y (1) to y (3) and synthesizing a new image at the resulting feature coordinates (Fig. 1C) A number of authors (Bregler Omohundro, 1995; Saul Jordan, 1997; Beymer Poggio, 1995) have previously shown how learning from examples allows sophisticated 3 The map from feature vectors to images was learned by fitting a GRBF net to 1000 corresponding points in both spaces. Each point corresponds to a node in the graph G used to ....

Bregler, C. & Omohundro, S. (1995). Nonlinear image interpolation using manifold learning. NIPS 7. MIT Press.

....Mixture Models Another natural extension is to consider mixtures of curved Gaussians. One of the most popular extensions of standard Gaussian models is to mixture models such as the mixture of Gaussians [Nowlan 1991] and in more recent work, mixtures of principal component or factor analyzers [Bregler and Omohundro 1995, Kambhatla and Leen 1997, Hinton et al. 1997, Roweis and Ghahramani 1999, Tipping 1997] Such mixture models are attractive because the expectation maximization algorithm (EM; Dempster et al. 1977] allows them to be fit to data in a computationally efficient manner. The curved Gaussian model can ....

Bregler, C. and S. M. Omohundro. 1995. Nonlinear image interpolation using manifold learning. In Tesauro, G., D. S. Touretzky, and T. K. Leen, editors, Advances in Neural and Information Processing Systems, 7, pages 971--980, Cambridge, MA. MIT Press.

....paradigm is to model nonlinear structure with a collection, or mixture, of local linear sub models. This philosophy is an attractive one, motivating, for example, the mixture of experts technique in regression problems [9] and the approach has indeed been adopted in the context of PCA [10, 3, 7, 6]. A local model implementation of PCA implies the integration of two procedures: a partitioning of the data space into distinct regions, and the estimation of the principal axes within each such region. The question of how to combine these two elements can be problematic as the optimal ....

C. Bregler and S. M. Omohundro. Nonlinear image interpolation using manifold learning. In G. Tesauro, D. S. Touretzky, and T. K. Leen, editors, Advances in Neural Information Processing Systems 7, pages 973--980. Cambridge, Mass: MIT Press, 1995.

....because the dimensionality reduction is geared to Gaussian mixture components. Many variants on local linear models have been proposed in the machine learning literature. Successful applications include handwritten digit recognition[10] data compression[13] and nonlinear image interpolation[3]. Factor analysis uses a small number of parameters to model the covariance structure of high dimensional data. For automatic speech recognition, these parameters can be chosen in two ways: i) to maximize the likelihood of observed speech signals, or (ii) to minimize the number of classification ....

Bregler, C. and Omohundro, S. (1995) Nonlinear image interpolation using manifold learning. In G. Tesauro, D. Touretzky, and T. Leen, eds. Advances in Neural Information Processing Systems 7:971--980. Cambridge: MIT Press.

....width and height, etc. to the controls of a polygonal lip model [1] Still others have a trained model of lip variations and attempt to fit the observations to this model. Some of the most interesting work done in this area has been along these lines: Bregler and Omohundro s work, for example [5], models the non linear subspace of valid lip poses within the image space and can thus be used for both analysis and synthesis. Similarly, Luettin s system learns the subspace of variations for 2D contours surrounding the lips [11] However, in order for these 2D models to be robust, they have to ....

Christoph Bregler and Stephen M. Omohundro. "Nonlinear Image Interpolation using Manifold Learning". In NIPS 7, 1995.

....width and height, etc. to the controls of a polygonal lip model [1] Still others have a trained model of lip variations and attempt to fit the observations to this model. Some of the most interesting work done in this area has been along these lines: Bregler and Omohundro s work, for example [5], models the non7 linear subspace of valid lip poses within the image space and can thus be used for both analysis and synthesis. Similarly, Luettin s system learns the subspace of variations for 2D contours surrounding the lips [11] However, in order for these 2D models to be robust, they have ....

Christoph Bregler and Stephen M. Omohundro. "Nonlinear Image Interpolation using Manifold Learning". In Advances in Neural Information Processing Systems 7, 1995.

....on image data [5] 7] others use such low level features to form a parametrized description of the lip shape [1] Some of the most interesting work done in this area has been in using a statistically trained model of lip variations. Bregler and Omohundro s work and Luettin s work, for example ([4] and [9] model the subspace of lip bitmaps and contours respectively. However, since these are 2D models, the changes in the apparent lip shape due to rigid rotations have to be modeled as complex changes in the lip pose. One goal of this paper is to extend these ideas to 3D. By modeling the ....

Christoph Bregler and Stephen M. Omohundro. "Nonlinear Image Interpolation using Manifold Learning". In NIPS 7, 1995.

....the same face, connect them by a smooth animation of intermediate poses[1] ii) given a telephone signal masked by intermittent noise, fill in the missing speech. Both these examples may be viewed as instances of the same abstract problem. In qualitative terms, we can state the problem as follows[2]: given a multidimensional data set, and two points from this set, find a smooth adjoining path that is consistent with available models of the data. We will refer to this as the problem of model based interpolation. In this paper, we examine the problem of non linear interpolation as it arises ....

....along these lines: in section 2, we consider the problem of intracluster interpolation; in section 3, the problem of intercluster interpolation. Previous approaches to non linear interpolation have exploited the properties of radial basis function (RBF) networks[1] and locally linear models[2]. We have been influenced by both these works, especially in the abstract formulation of the problem. Nevertheless, our approach has a number of distinguishing features namely, the fundamental role played by the density, the treatment of non Gaussian models, the use of a continuous variational ....

[Article contains additional citation context not shown here]

C. Bregler and S. Omohundro. Nonlinear image interpolation using manifold learning. In G. Tesauro, D. Touretzky, and T. Leen (eds.). Advances in Neural Information Processing Systems 7, 973--980. MIT Press, Cambridge, MA (1995).

No context found.

BREGLER, C., AND OMOHUNDRO, S. 1995. Nonlinear image interpolation using manifold learning. In Advances in Neural Information Processing Systems 7. 973--980.

No context found.

C. Bregler and S. Omohundro. Nonlinear image interpolation using manifold learning. In NIPS--7, 1995.

No context found.

Bregler, C. & Omohundro, S.M. (1995) Nonlinear image interpolation using manifold learning. In G. Tesauro, D.S. Touretzky & T.K. Leen (eds.), Advances in Neural Information Processing Systems 7: 973--980. MIT Press.

No context found.

C. Bregler and S. M. Omohundro, "Nonlinear image interpolation using manifold learning", in Advances in Neural Information Processing Systems 7,G.Tesauro, D. S. Touretzky, and T. K. Leen, Eds., pp. 971--980. MIT Press, 1995.

No context found.

C. Bregler and S. Omohundro. Nonlinear image interpolation using manifold learning. In NIPS--7, 1995.

No context found.

Bregler, C. & Omohundro, S.M. (1995) Nonlinear image interpolation using manifold learning. In G. Tesauro, D.S. Touretzky & T.K. Leen (eds.), Advances in Neural Information Processing Systems 7: 973--980. MIT Press.

No context found.

Bregler, C. & Omohundro, S. (1995) Nonlinear image interpolation using manifold learning. Advances in Neural Information Processing Systems 7. MIT Press.

No context found.

Bregler, C. & Omohundro, S.M. (1995) Nonlinear image interpolation using manifold learning. In G. Tesauro, D.S. Touretzky & T.K. Leen (eds.), Advances in Neural Information Processing Systems 7: 973--980. MIT Press.

No context found.

C. Bregler and S. Omohundro, Nonlinear image interpolation using manifold learning, in NIPS, vol. 7, 1995.

First 50 documents

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