Object detection is challenging partly due to the limited discriminative power of local feature descriptors. We amend this limitation by incorporating spatial constraints among ne...
A general approach for multiscale orientation analysis of 2D scalar images is proposed. A scale-dependent orientation bundle (map of the visual space into function of two arguments...
Stiliyan Kalitzin, Bart M. ter Haar Romeny, Max A....
In this work, we address the problem of building recognition across two camera views with large changes in scales and viewpoints. The main idea is to construct a semantically rich...
— This paper introduces a signal-recognition based approach for detecting autonomous mobile robot immobilization on outdoor terrain. The technique utilizes a support vector machi...
Object detection remains an important but challenging task in computer vision. We present a method that combines high accuracy with high efficiency. We adopt simplified forms of...