Keynote Lectures

In this talk, I will present an automatic system which dynamically tracks human objects and create their 3-D visualization from big visual data, collected either from a single or an array of static/moving cameras. These cameras are continuously calibrated among one another in a fully unsupervised manner so that the tracking across multiple cameras can be effectively integrated and reconstructed via the 3D open map service such as Google Earth or Microsoft Virtual Earth. Finally, I will also present some systematic recognition of the actions performed by the tracked human independent of the viewing perspectives.