2006:Audio Onset Detection Results: Solo Singing Voice

From MIREX Wiki

Solo Singing Voice Class - Peak F-measure Performance Across All Parameterizations

Contestant Parameters Total Correct Total FP Total FN Total Merged Total Doubled Avg. Correct Avg. FP Avg. FN Avg. Merged Avg. Doubled Avg. Precision Avg. Recall Avg. F-Measure
brossier_complex 0.5 95 196 134 1 10 31.7 65.3 44.7 0.3 3.3 0.372 0.417 0.387
brossier_dual 0.4 126 123 103 1 19 42.0 41.0 34.3 0.3 6.3 0.524 0.568 0.531
brossier_hfc 0.35 119 106 110 1 12 39.7 35.3 36.7 0.3 4.0 0.536 0.528 0.525
brossier_specdiff 0.4 52 200 177 0 0 17.3 66.7 59.0 0.0 0.0 0.289 0.241 0.255
dixon_cd (0.80/0.15) 107 1246 122 1 10 35.7 415.3 40.7 0.3 3.3 0.080 0.497 0.137
dixon_nwpd (0.89/0.60) 209 1750 20 2 49 69.7 583.3 6.7 0.7 16.3 0.122 0.917 0.214
dixon_rcd (0.88/0.40) 123 960 106 1 13 41.0 320.0 35.3 0.3 4.3 0.110 0.581 0.184
dixon_sf (0.83/0.35) 126 1032 103 1 12 42.0 344.0 34.3 0.3 4.0 0.107 0.581 0.178
dixon_wpd (0.84/0.65) 149 667 80 1 23 49.7 222.3 26.7 0.3 7.7 0.189 0.686 0.294
du 2.9 139 416 90 1 0 46.3 138.7 30.0 0.3 0.0 0.252 0.629 0.358
roebel_1 0.09 116 91 113 1 1 38.7 30.3 37.7 0.3 0.3 0.591 0.538 0.555
roebel_2 0.27 109 95 120 1 1 36.3 31.7 40.0 0.3 0.3 0.547 0.487 0.513
roebel_3 0.09 130 179 99 1 2 43.3 59.7 33.0 0.3 0.7 0.470 0.589 0.510

download these results as csv

Solo Singing Voice Class - Precision vs. Recall Curves

File:Solo singing voice all runs.png