Palestrante: Prof. Rodrigo de Bem, Centro de Ciências Computacionais da Universidade Federal do Rio Grande - FURG.

Horário / Local: 02:00, 26 de fevereiro de 2013

O prof. Rodrigo De Bem, professor do C3, dará uma palestra na Universidade de Surrey, UK, nesta terça-feira (26/02), falando sobre Rio Grande/Furg/C3/Nautec e do trabalho apresentado no Visapp 2013, evento que ocorreu na semana passada.

Abstract:

Despite the fact that 3D motion tracking has been highly explored in computer vision, it still faces some relevant challenges, such as the tracking of objects using little a priori knowledge. In this context, this work presents the Volume Geometric Decomposition method, which is capable of constructing representation models of distinct and previously unknown objects. This method is executed over a probabilistic volumetric reconstruction of the interest objects. It adjusts the representation to the reconstructed volume, minimizing the amount of empty space enclosed by the model. Such representation model is composed by an appearance and a kinematic model. The former comprises ellipsoids and joints, while the latter is implemented through the Loose-Limbed model, a probabilistic graphical model. We performed experiments on a 3D motion tracking framework, and the obtained results show that the proposed method successfully constructs models of highly distinct objects that were not know a priori.

Os detalhes da palesta estão nos links abaixo: 

http://www.surrey.ac.uk/cvssp/news/events/index.htm

http://www.surrey.ac.uk/cvssp/news/events/cvssp_and_ee/

3d_representation_models_construction_through_a_volume_geometric_decomposition_method.htm