T. Heap and D. Hogg. Automated pivot location for the cartesian-polar hybrid point distribution model. In R. Fisher and E. Trucco, editors, 7 Edinburgh, UK, Sept. 1996. BMVA Press.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....from those in the training set, and have more modes of variation than the true number of degrees of freedom needed to describe the objects variation. We have recently [2] described a polynomial regression PDM (PRPDM) which succeeds in capturing limited non linear shape variability. Heap and Hogg [3], on the other hand, describe using polar co ordinates for sub parts of a PDM, allowing them to rotate. Although potentially useful, this is not as general a model as the MLP based approach described here. In this paper we present a new method for modelling non linear shape variability which ....

Heap, T. and Hogg, D., 'Automated Pivot Location for the Cartesian-Polar Hybrid Point Distribution Model', 6th British Machine Vision Conference, Birmingham, 11-14 September, (1995).

....work worth noting as being relevant to the task of joint centre location. Closely related work is that of Ashbrook et al. [2, 3] who address the automatic construction of models of articulated objects from range data. Probably the most closely related work to our own is that of Heap and Hogg [7] who investigated pivot location for the study of articulated objects by point distribution models in polar co ordinates. They presented a least squares expression similar to the analytic method in this paper. Ours differs in two minor ways. We extend the 2D case to 3D, and byadifferentchoice of ....

T. Heap and D. Hogg. Automated pivot location for the cartesian-polar hybrid point distribution model. In British Machine Vision Conference, pages 97--106, Birmingham, England, 1995.

....A number of models have been proposed to deal with nonlinear shape variation. However, they often su er from certain drawbacks. Some involve a complicated model construction procedure [3] Some are supervised in the sense that they assume prior knowledge on the structure of the nonlinearity [12]. Others require prior classi cation with the number of classes to be estimated or speci ed beforehand and each class being assumed Gaussian [13, 5] And some cannot be easily extended to shape spaces of higher dimension [11] In the present paper we present a density estimation approach which is ....

T. Heap and D. Hogg. Automated pivot location for the cartesian-polar hybrid point distribution model. In BMVC, pages 97-106, Edinburgh, UK, Sept. 1996.

....both left and right hands. Shapes of di erent classes are morphed in an undesirable way. Several approaches have been undertaken to model nonlinear shape variability. They often su er from certain drawbacks, namely they assume some prior knowledge about the structure of the nonlinearity [8], or the number of underlying classes [3] or they involve an intricate model construction [2] An elegant and promising way to avoid these drawbacks is to employ feature spaces induced by Mercer kernels [1] in order to indirectly model a nonlinear transformation (x) of the original data from a ....

T. Heap and D. Hogg. Automated pivot location for the cartesian-polar hybrid point distribution model. In Brit. Machine Vision Conference, pages 97-106, Edinburgh, UK, Sept. 1996.

....of the object rotate, or there are changes in viewing position of a 3D object. There have been several non linear extensions to the PDM, either using polynomial modes [66] using a multi layer perceptron to perform non linear PCA [65] or using polar coordinates for rotating sub parts of the model [31]. However, all these approaches assume that varying the parameters b within given limits will always generate plausible shapes, and that all plausible shapes can be so generated. This is not always the case. For instance, if a sub part of the shape can appear in one of two positions, but not ....

....To generate new examples using the model which are similar to the training set we must constrain the parameters b to be near the edge of the circle. Points at the mean (b = 0) should actually be illegal. One approach would be to use an alternative parameterisation of the shapes. Heap and Hogg [31] use polar coordinates for some of the model points, relative to other points. A more general approach is to use non linear models of the probability density function, p(b) This allows the modelling of distinct classes of shape as well as non linear shape variation, and does not require any ....

T. Heap and D. Hogg. Automated pivot location for the cartesian-polar hybrid point distribution model. In R. Fisher and E. Trucco, editors, 7 th British Machine Vison Conference, pages 97-106, Edinburgh, UK, Sept. 1996. BMVA Press.

....of the object rotate, or there are changes in viewing position of a 3D object. There have been several non linear extensions to the PDM, either using polynomial modes [59] using a multi layer perceptron to perform non linear PCA [58] or using polar coordinates for rotating sub parts of the model [25]. However, all these approaches assume that varying the parameters b within given limits will always generate plausible shapes, and that all plausible shapes can be so generated. This is not always the case. For instance, if a sub part of the shape can appear in one of two positions, but not ....

....To generate new examples using the model which are similar to the training set we must constrain the parameters b to be near the edge of the circle. Points at the mean (b = 0) should actually be illegal. One approach would be to use an alternative parameterisation of the shapes. Heap and Hogg [25] use polar coordinates for some of the model points, relative to other points. A more general approach is to use non linear models of the probability density function, p(b) This allows the modelling of distinct classes of shape as well as non linear shape variation, and does not require any ....

T. Heap and D. Hogg. Automated pivot location for the cartesian-polar hybrid point distribution model. In R. Fisher and E. Trucco, editors, 7 th British Machine Vison Conference, pages 97-106, Edinburgh, UK, Sept. 1996. BMVA Press.

....of the object rotate, or there are changes in viewing position of a 3D object. There have been several non linear extensions to the PDM, either using polynomial modes [11] using a multi layer perceptron to perform non linear PCA [10] or using polar co ordinates for rotating sub parts of the model [5]. However, all these approaches assume that varying the parameters b within given limits will always generate plausible shapes, and that all plausible shapes can be so generated. This is not always the case. For instance, if a sub part of the shape can appear in one of two positions, but not ....

T. Heap and D. Hogg. Automated pivot location for the cartesian-polar hybrid point distribution model. In R. Fisher and E. Trucco, editors, 7 th British Machine Vison Conference, pages 97--106, Edinburgh, UK, Sept. 1996. BMVA Press.

....the training set using a backpropagation neural network to perform non linear principal component analysis. This performs well, however, the architecture of the network is application specific, and also the training times and the optimisation of network structure are time consuming. Heap and Hogg [6] suggested using a log polar mapping to remove non linearity from the training set. This allows a non linear training set to be projected into a linear space, where PCA can be used to represent deformation. The model is then projected back into the original space. Although a useful suggestion for ....

T. Heap, D.C. Hogg, Automated pivot location for the cartesian--polar hybrid point distribution model, in: D. Pycock (Ed.), British Machine Vision Conference 1995, British Machine Vision Association, Birmingham, UK, 1995, pp. 97--106.

....other work worth noting as being relevant to the task of joint centre location. Closely related work is that of Ashbrook et al. [2, 3] who address the automatic construction of models of articulated objects from range data. Probably the most closely related work to our own is that of Heap and Hogg [7] who investigated pivot location for the study of articulated objects by point distribution models in polar co ordinates. They presented a least squares expression similar to the analytic method in this paper. Ours di ers in two minor ways. We extend the 2D case to 3D, and by a di erent choice of ....

Tony Heap and D. Hogg. Automated pivot location for the cartesian-polar hybrid point distribution model. In British Machine Vision Conference, pages 97-106, Birmingham, England, 1995.

....kinematic knowledge a priori [14, 13, 9, 11, 5, 7] Apart from human tracking there is other work worth noting. Ashbrook et al. [2, 3] address the automatic construction of models of articulated objects from range data. Probably the most closely related work to our own is that of Heap and Hogg [6] who investigated pivot location for the study of articulated objects by point distribution models in polar co ordinates. 3 The problem We suppose that there are 2 objects segments labelled by = 0; 1 and each object has markers labelled by i = 0: N 1. The marker position is given in the ....

Tony Heap and D. Hogg. Automated pivot location for the cartesian-polar hybrid point distribution model. In British Machine Vision Conference, pages 97-106, Birmingham, England, 1995.

....other work worth noting as being relevant to the task of joint centre location. Closely related work is that of Ashbrook et al. [2, 3] who address the automatic construction of models of articulated objects from range data. Probably the most closely related work to our own is that of Heap and Hogg [7] who investigated pivot location for the study of articulated objects by point distribution models in polar co ordinates. They presented a least squares expression similar to the analytic method in this paper. Ours differs in two minor ways. We extend the 2D case to 3D, and by a different choice ....

T. Heap and D. Hogg. Automated pivot location for the cartesian-polar hybrid point distribution model. In British Machine Vision Conference, pages 97--106, Birmingham, England, 1995.

....from those in the training set, and have more modes of variation than the true number of degrees of freedom needed to describe the objects variation. We have recently [2] described a polynomial regression PDM (PRPDM) which succeeds in capturing limited non linear shape variability. Heap and Hogg [3], on the other hand, describe using polar co ordinates for sub parts of a PDM, allowing them to rotate. Although potentially useful, this is not as general a model as the MLP based approach described here. In this paper we present a new method for modelling non linear shape variability which ....

Heap, T. and Hogg, D., `Automated Pivot Location for the Cartesian-Polar Hybrid Point Distribution Model', 6th British Machine Vision Conference, Birmingham, 11-14 September, (1995).

No context found.

T. Heap and D. Hogg. Automated pivot location for the cartesian-polar hybrid point distribution model. In R. Fisher and E. Trucco, editors, 7 Edinburgh, UK, Sept. 1996. BMVA Press.

No context found.

Heap T, Hogg D. Automated Pivot Location for the Cartesian-Polar Hybrid Point Distribution Model. In: Pycock D, ed. British Machine Vision Conference

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