High Impact Factor : 4.396 icon | Submit Manuscript Online icon |

Song Search Engine Based On Querying By Singing/Humming

Author(s):

Shruti Mengade , AISSMS' Institute of Information Technology,Pune; Pritesh Patil, AISSMS Institute of Information Technology,Pune; Prapti Kolpe, AISSMS Institute of Information Technology,Pune; Vaishanvi Gawande, AISSMS Institute of Information Technology,Pune; Kalyani Budhe, AISSMS Institute of Information Technology,Pune

Keywords:

Query by humming/singing, Music Information Retrieval, query result re-ranking, edit distance, contour string generation

Abstract

Song Search Engine Based on Querying by Singing/Humming is the system which searches the song for user when given a query in the form of hummed or sung section of melody as a input to system. Searching song by singing /humming section of it is the most natural for user. For this no musical training would be required. User is simply expected to record the acoustic query and system will use the features of recorded query for searching song. Finally the query by singing/humming system will return optimized list of songs in which most probably top most songs will be the song desired by user. Result optimization is obtained by re-ranking algorithms using probabilistic calculation. Re-ranking of the result is advancement in the existing QBSH systems. This module has made the QBSH system more accurate & responsive.

Other Details

Paper ID: IJSRDV3I1026
Published in: Volume : 3, Issue : 1
Publication Date: 01/04/2015
Page(s): 14-16

Article Preview

Download Article