Difference between revisions of "2005:Audio Tempo Extraction Results"
From MIREX Wiki
| Line 1: | Line 1: | ||
==Introduction== | ==Introduction== | ||
| − | + | These are the results for the 2005 running of the Audio Tempo Extraction task. | |
| − | ===Dataset | + | ===Goal=== |
| + | The comparison and evaluation of current methods for the extraction of tempo from musical audio | ||
| + | |||
| + | ===Dataset=== | ||
| + | 140 wav files, 354 Megabytes | ||
==Results== | ==Results== | ||
Revision as of 10:21, 2 August 2010
Introduction
These are the results for the 2005 running of the Audio Tempo Extraction task.
Goal
The comparison and evaluation of current methods for the extraction of tempo from musical audio
Dataset
140 wav files, 354 Megabytes
Results
| Rank | Participant | Score (std. deviation) | At Least One Tempo Correct | Both Tempos Correct | At Least One Phase Correct | Both Phases Correct | Mean Absolute Difference of Scored Saliences | Runtime (s) | Machine |
|---|---|---|---|---|---|---|---|---|---|
| 1 | Alonso, David, & Richard | 0.689 (0.231) | 95.00% | 55.71% | 25.00% | 5.00% | 0.239 | 2875 | G |
| 2 | Uhle, C. (1) | 0.675 (0.273) | 90.71% | 59.29% | 32.14% | 7.14% | 0.222 | 1160 | F |
| 3 | Uhle, C. (2) | 0.675 (0.272) | 90.71% | 59.29% | 32.86% | 6.43% | 0.222 | 2621 | F |
| 4 | Gouyon & Dixon (1) | 0.670 (0.252) | 92.14% | 56.43% | 40.71% | 7.86% | 0.311 | 3303 | G |
| 5 | Peeters, G. | 0.656 (0.223) | 95.71% | 47.86% | 27.86% | 4.29% | 0.258 | 2159 | R |
| 6 | Gouyon & Dixon (2) | 0.649 (0.253) | 92.14% | 51.43% | 37.14% | 5.71% | 0.305 | 2050 | G |
| 7 | Gouyon & Dixon (4) | 0.645 (0.294) | 87.14% | 55.71% | 48.57% | 10.71% | 0.313 | 1357 | G |
| 8 | Eck, D. | 0.644 (0.300) | 86.43% | 53.57% | 37.14% | 5.71% | 0.230 | 1665 | Y |
| 9 | Davies & Brossier | 0.628 (0.284) | 86.43% | 48.57% | 26.43% | 4.29% | 0.224 | 1005 | R |
| 10 | Gouyon & Dixon (3) | 0.607 (0.287) | 87.14% | 47.14% | 36.43% | 6.43% | 0.294 | 1388 | R |
| 11 | Sethares, W. | 0.597 (0.252) | 90.71% | 37.86% | 30.71% | 0.71% | 0.239 | 70975 | Y |
| 12 | Brossier, P. | 0.583 (0.333) | 80.71% | 51.43% | 28.57% | 2.14% | 0.223 | 180 | B 0 |
| 13 | Tzanetakis, G. | 0.538 (0.359) | 71.43% | 50.71% | 28.57% | 3.57% | 0.295 | 7173 | B 0 |
McNemar's Test Results
Statistical probability that algorithms have same error function: Note: Results of less than 5% indicate significant differences, results of 1% or less indicate highly significant differences.
| Tzanetakis, G. | Gouyon & Dixon (3) | Gouyon & Dixon (1) | Gouyon & Dixon (0) | Brossier, P. | Uhle, C. (1) | Uhle, C. (0) | Sethares, B | Peeters, G | Eck, D. | Davies, M. | Alonso, David, & Richard | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Tzanetakis, G. | n/a | |||||||||||
| Gouyon & Dixon (3) | 17.44% | n/a | ||||||||||
| Gouyon & Dixon (1) | 50.00% | 19.58% | n/a | |||||||||
| Gouyon & Dixon (0) | 12.15% | 50.00% | 4.61% | n/a | ||||||||
| Brossier, P. | 50.00% | 23.54% | 55.31% | 20.05% | n/a | |||||||
| Uhle, C. (1) | 2.88% | 26.64% | 6.76% | 31.36% | 5.41% | n/a | ||||||
| Uhle, C. (0) | 2.88% | 26.64% | 6.31% | 30.89% | 5.86% | 75.00% | n/a | |||||
| Sethares, B. | 0.99% | 0.02% | 0.27% | 0.02% | 1.24% | 0.01% | 0.01% | n/a | ||||
| Peeters, G. | 34.70% | 7.19% | 29.83% | 6.32% | 30.73% | 2.40% | 2.40% | 2.97% | n/a | |||
| Eck, D. | 32.58% | 38.04% | 38.54% | 32.20% | 38.54% | 12.15% | 12.15% | 0.41% | 18.32% | n/a | ||
| Davies, M. | 38.77% | 7.17% | 33.59% | 6.31% | 33.89% | 2.67% | 2.67% | 1.38% | 50.00% | 21.35% | n/a | |
| Alsonso, David, & Richard | 18.02% | 56.12% | 23.99% | 50.00% | 23.99% | 26.64% | 26.64% | 0.04% | 7.62% | 39.39% | 6.07% | n/a |
