Excerpt-based summarization of popular music
|This page contains audio data for the
related paper "Automatic Music
Summarization via Similarity Analysis." This music is reproduced here under
"Fair Use" clause of the 1976 Copyright Act (17
|In this paper we use similarity analysis of digital audio to
select excerpts as summaries of a larger work. The excerpts are
selected to maximize average similarity to the entire piece.
Please refer to the paper for more
Example test signal. This 100-second audio file
consists of 30
seconds of a 1 kHz sine wave, 40 seconds of a 500 Hz sine wave, and 30
of a 2 kHz sine wave. Because the 500 Hz portion is the longest, the
summary should consist primarily of the 500 Hz signal as opposed to the
1 KHz and 2 KHz segments.
N.B.: These files contain nearly two minutes of highly
tones, and there's no compelling reason for anyone to actually listen
We provide them for completeness, and for anyone who might care to
results using this data.
Table 1: Summary times for The Magical Mystery Tour,
Table 2: Summary times for Spring - Allegro,
Four Seasons, by Antonio Vivaldi:
Table 3: Summary times for Wild Honey, by U2
For comparison, 30 second summary from Amazon.com: Wild
Table 4: Summary times for Take the %u201CA%u201D Train,
Ellington and his orchestra.
| M. Cooper, and
J. Foote. Automatic Music Summarization via Similarity Analysis. Proc.
Intl. Symposium on Music Information Retrieval, pp. 81-5, 2002. [PDF]
Technical Contact: Matt Cooper