Using Melody Matching Model Based on a String Pattern Matching Mechanism of Frequency Contours

Abstract

Music is a universal language. Music originated and evolved with developing demands of the human and the rapid discoveries in science and technology. Sharing music over World Wide Web is an upcoming trend among computer users. Simply, the music database owners are using the web to fulfill the needs of the people who are fed up of searching for their favorite music over different methods. Sometimes those music search engines provide enough facility to find the piece of music that the user wants but sometimes not. Because it is hard to implement a music search engine which can provide all types of searching techniques that users like to have, it makes it inconvenient for music lovers to search music by reminding metadata of a song such as song title, author name, lyrics, etc. Melody can be considered as the most captivating and memorable feature of a song because the human brain can easily capture the melody rather than remembering metadata of a song. Therefore, Query by Humming (QBH) refers to music information retrieval systems where short audio clips of singing or humming act as queries. Melody is considered as the most important feature in the queries and the songs. This system proposes a QBH system Using Melody Matching Model based on a string pattern matching mechanism of frequency contours, which is used to align the input pitch frequency contour with the original music pitch frequency contour.