We present a model-based approach to recover the rigid and non-rigid facial motion parameters in video sequences. Our face model is based on anatomically motivated muscle actuator controls to model the articulated non-rigid motion of a human face. The model is capable of generating a variety of facial expressions by using a small number of muscle actuator controls. We estimate rigid and non-rig...

A new method is presented for computing the fundamental matrix from point correspondences: its singular value decomposition (SVD) is optimized by the Levenberg-Marquard (LM) method. The search is initialized by optimal correction of unconstrained ML. There is no need for tentative 3-D reconstruction. The accuracy achieves the theoretical bound (the KCR lower bound).

The present work is aimed at modelling and characterizing viscoelastic materials by means of a time domain technique. A constitutive equation containing internal variables is proposed. The parameter estimation problem is solved in time domain by means of the Levenberg-Marquadt technique. The effectiveness of the time domain estimation is assessed experimentally for three different structres.

A new method for the solution of the generalized complementarily problem is introduced. The method is based on a no smooth equation reformulation of the generalized complementarily problem and on a no smooth Levenberg-Marquardt method for its solution. The method is shown to be globally convergent. Numerical results are also given.

A new, to my knowledge, procedure for retrieving the wave aberration from the point-spread function is presented. It uses the Levenberg-Marquardt optimization algorithm in a mutiresolution pyramidal scheme. The method, tested with simulated large aberrations without initial estimates, accelerates convergence and avoids stagnation in local minima.

*This research was carried out as part of NWO research project 611-304-019, 'Address: Free University, Department of Economics and Econometrics, De Boelelaan 1105, 1081 HV Amsterdam, The Netherlands. E-mail: [email protected].

In this paper, the problem of nonnegative matrix factorization (NMF) is considered. It is formulated as the optimization of a criterion with bound constraints. We propose an approach based on Givens parameterization of some positive vector, and criterion minimization is achieved using Levenberg-Marquardt algorithm. The performance of the developed NMF method is illustrated for the separation of...

This work approaches the problem of estimating Kautz optimal poles in kernel expansion Wiener-Volterra models. The analytical solution for suboptimal case is already established literature. However, two parameters that compose still open. In this paper, an optimization strategy using Levenberg-Marquardt presented. algorithm used to find parameters, with same base all dimensions. construction ba...

We classify and review existing algorithms for computing the fundamental matrix from point correspondences and propose new effective schemes: 7-parameter Levenberg-Marquardt (LM) search, EFNS, and EFNS-based bundle adjustment. Doing experimental comparison, we show that EFNS and the 7-parameter LM search exhibit the best performance and that additional bundle adjustment does not increase the ac...