Sergio Hernandez, PhD Student

Biography

Sergio studied acoustic engineering at the Universidad Tecnologica de Chile where he did his undergraduate thesis in the subject "Optimization model for the stockbridge vibration neutralizer", where an eigenvalue dimensionality reduction was used for an optimization problem in modal vibration analysis. Then, he completed a Masters in computer engineering at the Universidad de Santiago de Chile, where he did his thesis in the subject "Neuro-fuzzy method for defect detection in aluminum castings", where a neuro-fuzzy method for dimensionality reduction and classification was proposed for performing non destructive testings in aluminium castings images.

He is now working under the supervision of Marcus Frean and Paul Teal, in a method for tracking multiple objects in video sequences. The method uses random sets and point process theory under the Bayesian filtering framework, where the goal is to estimate the state of a model with time varying dimensions.