Difference between revisions of "2006:Audio Tempo Extraction Results"
IMIRSELBot (talk | contribs) m (Robot: Automated text replacement (-\[\[([A-Z][^:]+)\]\] +2006:\1)) |
|||
| Line 1: | Line 1: | ||
[[Category: Results]] | [[Category: Results]] | ||
==Introduction== | ==Introduction== | ||
| − | These are the results for the 2006 running of the Audio Tempo Extraction task set. For background information about this task set please refer to the [[Audio Tempo Extraction]] page. | + | These are the results for the 2006 running of the Audio Tempo Extraction task set. For background information about this task set please refer to the [[2006:Audio Tempo Extraction]] page. |
The aim of the Audio Tempo Extraction task is to extract the perceptual tempi from audio recordings. We differentiate the notion of perceived tempo and notated tempo in this task. A test set of 140 recordings was used to evaluate the algorithms. Ground truths for each recording were extracted from multiple listeners. Algorithms were scored on their ability to extract the two dominant perceptual tempi from a piece of music, with each tempo in the ground truth given a perceptual salience from the listening tests. In this task, a P-score of 1.0 denotes perfect performance. | The aim of the Audio Tempo Extraction task is to extract the perceptual tempi from audio recordings. We differentiate the notion of perceived tempo and notated tempo in this task. A test set of 140 recordings was used to evaluate the algorithms. Ground truths for each recording were extracted from multiple listeners. Algorithms were scored on their ability to extract the two dominant perceptual tempi from a piece of music, with each tempo in the ground truth given a perceptual salience from the listening tests. In this task, a P-score of 1.0 denotes perfect performance. | ||
Revision as of 13:04, 13 May 2010
Contents
Introduction
These are the results for the 2006 running of the Audio Tempo Extraction task set. For background information about this task set please refer to the 2006:Audio Tempo Extraction page.
The aim of the Audio Tempo Extraction task is to extract the perceptual tempi from audio recordings. We differentiate the notion of perceived tempo and notated tempo in this task. A test set of 140 recordings was used to evaluate the algorithms. Ground truths for each recording were extracted from multiple listeners. Algorithms were scored on their ability to extract the two dominant perceptual tempi from a piece of music, with each tempo in the ground truth given a perceptual salience from the listening tests. In this task, a P-score of 1.0 denotes perfect performance.
General Legend
Team ID
klapuri = Anssi Klapuri
davies = M. E. P. Davies and M. D. Plumbley
alonso = Miguel Alonso, Bertrand David and Gael Richard
ellis = Daniel P. W. Ellis
antonopoulos = Iasonas Antonopoulos, Aggelos Pikrakis and Sergios Theodoridis
brossier = Paul Brossier
Overall Summary Results
MIREX 2006 Audio Tempo Extraction Summary Results
file /nema-raid/www/mirex/results/at06_results.csv not found
MIREX 2006 Audio Tempo Extraction Runtime Data
file /nema-raid/www/mirex/results/at06_runtime.csv not found
Summary, plots, and comments on the results
Click here to get a summary, plots and comments on the contest.
