Since the initials of Artificial Intelligence, many learning techniques have been proposed to deal with many artificial systems. The initial learning designs were proposed to deal with just two classes. Which option is the best one given two previous possibilities? To solve this problem using several examples from two hypotheses, many learning techniques have been developed with successful results. However, in many real applications, it is common to face with problems where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$N>2$$\end{document} possible solutions exist. Multi-class classification is the term applied to those machine learning problems that require assigning labels to instances where the labels are drawn from a set of at least three classes. Once we have detected a traffic sign, we are in this kind of multi-class scenarios, where we have to discriminate among a set of sign categories. In this chapter, we review state-of-the-art multi-class methodologies based on the Error-Correcting Output Codes framework (ECOC). In particular, we describe different ECOC strategies for extending any kind of classifier to deal the problem of multiclass traffic sign classification.
Traffic Sign Categorization
SpringerBriefs Computer Sci.
2011-09-22
28 pages
Article/Chapter (Book)
Electronic Resource
English
TRAFFIC SIGN RECOGNITION DEVICE AND TRAFFIC SIGN RECOGNITION METHOD
European Patent Office | 2023
|TRAFFIC SIGN RECOGNITION DEVICE AND TRAFFIC SIGN RECOGNITION METHOD
European Patent Office | 2017
|