Authors: Bob Sturm
Type: Conference paper
Conference: 2012 IEEE Conference on Multimedia & Expo
Abstract: We explore risk and rejection for music genre recognition (MGR) within the minimum risk framework of Bayesian classification. In this way, we attempt to give an MGR system knowledge that some misclassifications are worse than others, and that deferring classification to an expert may be a better option than forcing a label under high uncertainty. Our experiments show this approach to have some success with respect to reducing false positives and negatives.