In this paper we present a model-based algorithm for the estimation of 3D motion and the analysis of facial expressions of a speaking person. A set of facial animation parameters ...
In this paper we propose a new method for human action categorization by using an effective combination of a new 3D gradient descriptor with an optic flow descriptor, to represent...
We believe it is important for web graphic standards such as SVG to support user interaction and diagrams that can adapt their layout and appearance to their viewing context so as...
Many problems require minimally perturbing an initial state in order to repair some violated constraints. We consider two search spaces for exactly solving this minimal perturbatio...
This paper presents a modulation-based reconstruction method for audio signals across long gaps of missing samples. We use LTI filterbanks followed by a multiplicative model that...