This paper presents a robust and accurate method for joint head pose and facial actions tracking, even under challenging conditions such as varying lighting, large head movements,...
Visual tracking is a challenging problem, as an object may change its appearance due to viewpoint variations, illumination changes, and occlusion. Also, an object may leave the fie...
— We present a self-calibrating photogeometric method using only off-the-shelf hardware that enables quickly and robustly obtaining multi-million point-sampled and colored models...
Missing data methods attempt to improve robust speech recognition by distinguishing between reliable and unreliable data in the time-frequency domain. Such methods require a binar...
—This paper presents a methodology to evaluate and optimize the robustness of an embedded system in terms of invariability in case of design revisions. Early decisions in embedde...