Dariush Alimohammadi, Keyvan Borna,
Volume 4, Issue 4 (3-2018)
Abstract
Background and Aim: The current research aims at prototyping query-by-humming music information retrieval systems for smart phones.
Methods: This multi-method research follows simulation technique from mixed models of the operations research methodology, and the documentary research method, simultaneously. Two chromatic harmonica albums comprised the research population. To achieve the purpose of research, 24 homophonic tracks were splitted by using Helium Audio Splitter software. The splits were processed by Sonic Visualiser software; and 168 XML documents were produced. On the other hand, 4 research participants hummed and recorded splits. Hummed tracks were converted by using AMR to MP3 Converter software, processed by Sonic Visualiser, and resulted in 672 XML documents. MATLAB software was learned by the first group of XML documents (168), and then, processed the second group of XML documents (672) for providing desirable outputs. Outputs were compared by using Image Comparer software.
Results: Findings indicated a high degree of similarity (99 %) between outputs of two groups of XML documents. It has also been found that the gender and the music skill do not have any impact on the results.
Conclusion: It could be acknowledged that designing query-by-humming systems based on converting audio to XML documents, and document matching, is an appropriate strategy towards developing music retrieval applications for smart phones.
