Difference between revisions of "2017:Automatic Lyrics-to-Audio Alignment Results"

From MIREX Wiki
(General Legend)
(Results)
Line 41: Line 41:
  
 
===Hansen's Dataset===
 
===Hansen's Dataset===
asda
+
AK1
 +
<csv>2017/ala/results_htk_hansen.csv</csv>
 
===Mauchs's Dataset===
 
===Mauchs's Dataset===

Revision as of 05:33, 16 October 2017

General Legend

Sub code Submission name Abstract Contributors
AK1 HMM-based Alignment PDF Anna Kruspe
AK2 DTW-based Alignment PDF Anna Kruspe
AK3 Levenshtein-based Alignment PDF Anna Kruspe
DMS1 Automatic Lyrics-to-Audio Alignment with Harmonic Modeling PDF Georgi Dzhambazov, Xavier Serra
DMS2 Automatic Lyrics-to-Audio Alignment with Source Separation PDF Georgi Dzhambazov, Marius Miron


Results

Hansen's Dataset A cappella

Summary Results

file /nema-raid/www/mirex/results/2017/ala/summary_results.csv not found

Per-song results

Hansen's Dataset

AK1

Track Average absolute error Percentage of correct segments
clocks 42.760 0.002
i_kissed_a_girl 7.640 0.052
rehab 6.040 0.061
rolling_in_the_deep 2.990 0.245
sunrise 3.240 0.308
this_afternoon 2.650 0.277
trick_me 18.130 0.078
umbrella 16.330 0.014
viva_la_vida 25.820 0.009

download these results as csv

Mauchs's Dataset