2009:Audio Beat Tracking
The aim of the automatic beat tracking task is to track each beat locations in a collection of sound files. Unlike the Audio Tempo Extraction task, which aim is to detect tempi for each file, the beat tracking task aims at detecting all beat locations in recordings. The algorithms will be evaluated in terms of their accuracy in predicting beat locations annotated by a group of listeners.
We have an additional data set this year of Chopin Mazurka pieces. There are on the order of 367 pieces with marked up beat locations. Because the tempo varies so much in these pieces, we expect it to be extremely difficult. Nevertheless, it should prove to be interesting for evaluations. We might select a subset of these 367 for the eval database.
The original old dataset are the same 160 30-second excerpts (WAV format) used for the Audio Tempo and Beat contests in 2006. Beat locations have been annotated in each excerpt by 40 different listeners (39 listeners for a few excerpts. The length of each excerpt is 30 seconds.
The older dataset audio recordings were selected to provide a stable tempo value, a wide distribution of tempi values, and a large variety of instrumentation and musical styles. About 20% of the files contain non-binary meters, and a small number of examples contain changing meters. One disadvantage of using this set for beat tracking is that the tempi are rather stable and this set will not test beat-tracking algorithms in their ability to track tempo changes.
The second mazurka dataset contains tempo changes so it will evaluate the ability of algorithms to track these.
Submitted programs should output one beat location per line, with a ┬½new line┬╗ character (\n) at the end of each line. The results should either be saved to a text file.
Example of possible output:
Each submission should be accompanied with a README file describing how the program should be used. For instance:
To run the program foobar on the file input.wav and store the results in the file output.txt, the following commands are examples of what should be used:
foobar -i input.wav > output.txt foobar -i input.wav -o output.txt foobar input.wav output.txt
- Matthew Davies (Queen Mary, University of London), <matthew.davies at elec.qmul.ac.uk>
- Douglas Eck (University of Montreal), <eckdoug at iro.umontreal.ca>
- Simon Dixon (Queen Mary, University of London) <simon.dixon at elec.qmul.ac.uk>
- Geoffroy Peeters (Ircam - CNRS) <geoffroy.peeters at ircam.fr>
- Tsung-Chi Lee (National Tsing Hua University of Taiwan), <leetc at mirlab.org>
This is a major re-write by Martin McKinney and is open to suggestions.
Evaluation of beat-tracking includes an implicit evaluation of tempo accuracy, however, the focus here will be on proper time position of beats. We propose the following evaluation method, which is quite simple in nature and accounts for ambiguity in the perception of the most salient metrical level: For each excerpt, an impulse train will be created from each of the 40 annotated ground truth beat vectors as well as from the algorithm output. The impulse trains will be 25 seconds long (ignoring tapped beats at times less than 5 seconds), constructed with a 100-Hz sampling rate, and have unit impulses at beat times. Each impulse train of annotations will be denoted by , where the subscript is the annotator number (1-40), and the impulse train from the algorithm will be donoted by . The performance, , of the beat-tracking algorithm for a single excperpt will be measured by calculating the cross-correlation function of and within a small delay window, , around zero and then averaged across the number of annotators ():
where is the sample-length of pulse trains and , and NP is a normalization factor defined by the maximum number of impulses in either impulse train:
The "error" window, W, is proportional to (1/5 of) the beat in the annotated taps and is defined (in Matlab notation ;-) as:
= round(0.2 * median(diff(find((a_s[n])))).
The algorithm with the best average P-score (across excerpts) will win.
The choice of 1/5 of the beat was somewhat arbitrarily chosen and is open for discussion. I've used this method to examine correlations between taps of different subjects and it works quite well. Comments please. -Martin
A collection of 160 musical exerpts will be used for the evaluation procedure, the same collection used for the 2006 Audio Tempo Extraction and Beat tasks. Each recording has been annotated by 40 different listeners (39 in a few cases). The annotation procedures were described in  and .
20 excerpts will be provided for training to the participant, and the remaining 140 excerpts, novel to all participants, will be used for the contest.
The second dataset consists of 367 mazurka performances, although a subset of these may be used.
- Masataka Goto and Yoichi Muraoka. Issues in evaluating beat tracking systems. In Working Notes of IJCAI-97 Workshop on Issues in AI and Music - Evaluation and Assessment, pages 9┬¡16, 1997 postscript
- McKinney, M.F. and Moelants, D. (2004), Deviations from the resonance theory of tempo induction, Conference on Interdisciplinary Musicology, Graz. pdf
- Moelants, D. and McKinney, M.F. (2004), Tempo perception and musical content: What makes a piece slow, fast, or temporally ambiguous? International Conference on Music Perception & Cognition, Evanston, IL. pdf
Leave comments here
You can find it here:
User: beattrack Password: b34trx
User: tempo Password: t3mp0
Data has been uploaded in both .tgz and .zip format.