Download This PaperAn Optimized Music Recognition System Using Mel-Frequency Cepstral Coefficient (MFCC) and Vector Quantization (VQ)
Research Directions: Special Issue International Business Research Conference on Transformation Opportunities and Sustainability Challenges in Technology and Management. UGC Journal No. 45489, UGC Sr. No. 1208, Impact Factor - 5.11 (UIF), pp 100 - 106, ISSN No. 2321 - 5488, Research Directions.
7 Pages Posted: 26 May 2020
Date Written: May 23, 2019
Abstract
At the present time large amounts of digital music began to be distributed through the internet and for which the demand for music related content like lyrics, musician profiles and movie clips has been increasing. The music related content has been offered through online services that usually use text based technology. This technology recognizes identical content by using file name or tag information associated with the music. Recognizing identical music content is difficult if file name or tag information is wrong. We therefore need a method for recognizing identical music content that uses audio data itself, in order to overcome these problems. In this paper, we propose a feature based identical music content recognition method for music related content services. We extracted feature data that have audible characteristics from waveform data of music content and processed it using a simplified version of Mel Frequency Cepstral Coefficient (MFCC) and Vector Quantization (VQ). Using this method we could identify consistent feature data, even if two wave forms used different digitizing specifications and were not exactly the same. The recognition method could find identical music content regardless of wave form changes. Our experiments prove that a simplified version of MFCC is effective in recognizing identical music content.
Keywords: Music Recognition, Mel - Frequency Cepstral Coefficient, Vector Quantization
JEL Classification: C61
Suggested Citation: Suggested Citation