Home // MMEDIA 2016, The Eighth International Conferences on Advances in Multimedia // View article

Two Kinds of Audio Feature Selection Algorithms in Wavelet Domain

Authors:
De Li
DaYou Jiang
Jongweon Kim

Keywords: Audio feature selection; DWT; DCT; MFCCs

Abstract:
In this paper, we propose two new audio feature selection algorithms based on Discrete Wavelet Transform (DWT) domain. One of them is combined with Discrete Cosine Transform (DCT), the other one is combined with Mel Frequency Cepstral Coefficients (MFCCs). First, we introduce two different audio selection algorithms; second, we use the audio attack experiments to verify reliability of those algorithms. The tests show that the DWT-DCT algorithm has better stability under most of audio attacks, but not is robust to amplify attack for electronic music, while DWT-MFCCs algorithm is especially stable under amplify attack, but not is robust to mp3 compression attacks.

Pages: 29 to 33

Copyright: Copyright (c) IARIA, 2016

Publication date: February 21, 2016

Published in: conference

ISSN: 2308-4448

ISBN: 978-1-61208-452-7

Location: Lisbon, Portugal

Dates: from February 21, 2016 to February 25, 2016