In this paper a method for shape estimation and structure extension of a surface using information from specularities is proposed. The structure of the scene is obtained from an i...
This paper presents a generative model based approach to man-made structure detection in 2D natural images. The proposed approach uses a causal multiscale random field suggested i...
Traditional approaches to Multiple-Instance Learning (MIL) operate under the assumption that the instances of a bag are generated independently, and therefore typically learn an in...
Abstract—We present a silicon neuron with a dynamic, active leak that enables precise spike-timing with respect to a time-varying input signal. Our neuron models the mammalian bu...
Changing conditions for teaching increase our motivation to understand the teaching and learning process. First time investigators of educational settings often feel uncertain abo...