EURASIP Journal on Advances in Signal Processing
Volume 2007 (2007), Article ID 56561, 15 pages
doi:10.1155/2007/56561
Research Article
Audio Key Finding: Considerations in System Design and Case Studies on Chopin's 24 Preludes
1Integrated Media Systems Center, Department of Computer Science, USC Viterbi School of Engineering, University of Southern California, Los Angeles 90089-0781, CA, USA
2Integrated Media Systems Center, Epstein Department of Industrial and Systems Engineering, USC Viterbi School of Engineering, University of Southern California, Los Angeles 90089-0193, CA, USA
Received 8 December 2005; Revised 31 May 2006; Accepted 22 June 2006
Academic Editor: George Tzanetakis
Copyright © 2007 Ching-Hua Chuan and Elaine Chew. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Abstract
We systematically analyze audio key finding to determine factors
important to system design, and the selection and evaluation of
solutions. First, we present a basic system, fuzzy analysis spiral
array center of effect generator algorithm, with three key
determination policies: nearest-neighbor (NN), relative distance
(RD), and average distance (AD). AD achieved a 79% accuracy
rate in an evaluation on 410 classical pieces, more than 8%
higher RD and NN. We show why audio key finding sometimes
outperforms symbolic key finding. We next propose three extensions
to the basic key finding system—the modified spiral array (mSA),
fundamental frequency identification (F0), and post-weight
balancing (PWB)—to improve performance, with evaluations using
Chopin's Preludes (Romantic repertoire was the most challenging).
F0 provided the greatest improvement in the first 8 seconds, while
mSA gave the best performance after 8 seconds. Case studies
examine when all systems were correct, or all incorrect.