Metric Learning for Music Symbol Recognition

• Type: Article in International Conference Proceedings Book
• Date: 2011

Authors

Rebelo, A (Author) Other; Tkaczuk, J (Author) Other; Sousa, R (Author) Other; Cardoso, JS (Author) FEUP

Conference proceedings International

Title: Proceedings - 10th International Conference on Machine Learning and Applications, ICMLA 2011

Pages: 106-111

10th International Conference on Machine Learning and Applications, ICMLA 2011

Honolulu, HI, 18 December 2011 through 21 December 2011

Indexing

ISI Web of Knowledge

Scopus - 7 Citations

Other information

Authenticus ID: P-008-2WP

DOI: 10.1109/ICMLA.2011.94

Abstract (EN): Although Optical Music Recognition (OMR) has been the focus of much research for decades, the processing of handwritten musical scores is not yet satisfactory. The efforts made to find robust symbol representations and learning methodologies have not found a similar quality in the learning of the dissimilarity concept. Simple Euclidean distances are often used to measure dissimilarity between different examples. However, such distances do not necessarily yield the best performance. In this paper, we propose to learn the best distance for the k-nearest neighbor (k-NN) classifier. The distance concept will be tuned both for the application domain and the adopted representation for the music symbols. The performance of the method is compared with the support vector machine (SVM) classifier using both real and synthetic music scores. The synthetic database includes four types of deformations inducing variability in the printed musical symbols which exist in handwritten music sheets. The work presented here can open new research paths towards a novel automatic musical symbols recognition module for handwritten scores. © 2011 IEEE.

Language: English

Type (Professor's evaluation): Scientific