Physiological modelling for improved reliability in silhouette-driven gradient-based hand tracking

We present a gradient-based motion capture system that robustly tracks a human hand, based on abstracted visual information - silhouettes. Despite the ambiguity in the visual data and despite the vulnerability of gradient-based methods in the face of such ambiguity, we minimise problems related to misfit by using a model of the hand's physiology, which is entirely non-visual, subject-invariant, and assumed to be known a priori. By modelling seven distinct aspects of the hand's physiology we derive prior densities which are incorporated into the tracking system within a Bayesian framework. We demonstrate how the posterior is formed, and how our formulation leads to the extraction of the maximum a posteriori estimate using a gradient-based search. Our results demonstrate an enormous improvement in tracking precision and reliability, while also achieving near real-time performance. © 2009 IEEE.

A presumed joint pdf model for turbulent combustion with varying equivalence ratio

Modeling of the joint probability density function of the mixture fraction and progress variable with a given covariance value is studied. This modeling is validated using experimental and direct numerical simulation (DNS) data. A very good agreement with experimental data of turbulent stratified flames and DNS data of a lifted hydrogen jet flame is obtained. The effect of using this joint pdf modeling to calculate the mean reaction rate with a flamelet closure in Reynolds averaged Navier-Stokes (RANS) calculation of stratified flames is studied. The covariance effect is observed to be large within the flame brush. The results obtained from RANS calculations using this modeling for stratified jet- and rod-stabilized V-flames are discussed and compared to the measurements as a posteriori validation for the joint probability density function model with the flamelet closure. The agreement between the computed and measured values of flame and turbulence quantities is found to be good. © 2012 Copyright Taylor and Francis Group, LLC.

Two problems with variational expectation maximisation for time-series models

Variational methods are a key component of the approximate inference and learning toolbox. These methods fill an important middle ground, retaining distributional information about uncertainty in latent variables, unlike maximum a posteriori methods (MAP), and yet generally requiring less computational time than Monte Carlo Markov Chain methods. In particular the variational Expectation Maximisation (vEM) and variational Bayes algorithms, both involving variational optimisation of a free-energy, are widely used in time-series modelling. Here, we investigate the success of vEM in simple probabilistic time-series models. First we consider the inference step of vEM, and show that a consequence of the well-known compactness property of variational inference is a failure to propagate uncertainty in time, thus limiting the usefulness of the retained distributional information. In particular, the uncertainty may appear to be smallest precisely when the approximation is poorest. Second, we consider parameter learning and analytically reveal systematic biases in the parameters found by vEM. Surprisingly, simpler variational approximations (such a mean-field) can lead to less bias than more complicated structured approximations.

Convex relaxation of mixture regression with efficient algorithms

We develop a convex relaxation of maximum a posteriori estimation of a mixture of regression models. Although our relaxation involves a semidefinite matrix variable, we reformulate the problem to eliminate the need for general semidefinite programming. In particular, we provide two reformulations that admit fast algorithms. The first is a max-min spectral reformulation exploiting quasi-Newton descent. The second is a min-min reformulation consisting of fast alternating steps of closed-form updates. We evaluate the methods against Expectation-Maximization in a real problem of motion segmentation from video data.

Interactive multi-objective particle swarm optimisation using decision space interaction

The most common approach to decision making in multi-objective optimisation with metaheuristics is a posteriori preference articulation. Increased model complexity and a gradual increase of optimisation problems with three or more objectives have revived an interest in progressively interactive decision making, where a human decision maker interacts with the algorithm at regular intervals. This paper presents an interactive approach to multi-objective particle swarm optimisation (MOPSO) using a novel technique to preference articulation based on decision space interaction and visual preference articulation. The approach is tested on a 2D aerofoil design case study and comparisons are drawn to non-interactive MOPSO. © 2013 IEEE.