A framework for music-speech segregation using music fingerprinting and acoustic echo cancellation principle. | [Nucleus • 2015]

Author(s):

1. F. Hussain: Department of Computer Engineering, University of Engineering and Technology Taxila, Taxila, Pakistan

2. H. A. Habib: Department of Computer Engineering, University of Engineering and Technology Taxila, Taxila, Pakistan

3. M. J. Khan: Department of Computer Engineering, University of Engineering and Technology Taxila, Taxila, Pakistan

Abstract:

Background interference creates voice intelligibility issue for listener. This research work considers background music as interference for communication through smart phone in areas with loud background music. This paper proposes a novel framework for background music segregation from human speech using music fingerprinting and acoustic echo cancellation. Initially, background music is searched in the database by music fingerprinting. Identified background music is registered and segregated using acoustic echo cancellation. Proposed approach generates better quality music speech segregation than existing algorithms. The research work is novel and segregates background music completely in comparison to existing approaches where single instruments are segregated successfully.

Page(s): 29-39

DOI: DOI not available

Published: Journal: Nucleus, Volume: 52, Issue: 1, Year: 2015

Keywords:

Keywords are not available for this article.

References:

References are not available for this document.

Citations

Citations are not available for this document.

Citations

Downloads

Views