A novel method for extracting parametric junction and corner features in images is presented. By treating each complex feature as a combination of elementary line and edge feature...
For the tasks of classification, two types of patterns can generate problems: ambiguous patterns and outliers. Furthermore, it is possible to separate classification algorithms in...
Jonathan Milgram, Mohamed Cheriet, Robert Sabourin
In this paper we present a combined approach for ob-
ject localization and classification. Our contribution is two-
fold. (a) A contextual combination of localization and clas-
...
We propose an algorithm for the classification of fluorescence microscopy images depicting the spatial distribution of proteins within the cell. The problem is at the forefront of...
Thomas E. Merryman, Keridon Williams, Gowri Sriniv...
We present a two-step method to speed-up object detection systems in computer vision that use Support Vector Machines (SVMs) as classifiers. In a first step we perform feature red...
Bernd Heisele, Thomas Serre, Sayan Mukherjee, Toma...