Difference between revisions of "2005:Audio Artist Identification Results"

From MIREX Wiki
Line 1: Line 1:
 
==Introduction==
 
==Introduction==
  
==Goal==
+
===Goal===
 
'''Goal:''' To identify artist from music audio (in PCM format).
 
'''Goal:''' To identify artist from music audio (in PCM format).
  
==Dataset==
+
===Datasets===
 
'''Dataset:''' Two sets of data were used: Magnatune and USPOP. The audio sampling rates used were either 44.1 KHz or 22.05 KHz (mono). More data information is in the following table.
 
'''Dataset:''' Two sets of data were used: Magnatune and USPOP. The audio sampling rates used were either 44.1 KHz or 22.05 KHz (mono). More data information is in the following table.
  
Line 18: Line 18:
 
|}
 
|}
  
===Result===
+
==Results==
  
==Overall==
+
===Overall===
 
{| border="1" cellspacing="0"
 
{| border="1" cellspacing="0"
 
|- style="background: yellow; text-align: center;"
 
|- style="background: yellow; text-align: center;"
Line 44: Line 44:
  
  
==Magnatune Dataset==
+
===Magnatune Dataset===
 
{| border="1" cellspacing="0"
 
{| border="1" cellspacing="0"
 
|- style="background: yellow; text-align: center;"
 
|- style="background: yellow; text-align: center;"
Line 73: Line 73:
 
|}
 
|}
  
==USPOP Dataset==
+
===USPOP Dataset===
 
{| border="1" cellspacing="0"
 
{| border="1" cellspacing="0"
 
|- style="background: yellow; text-align: center;"
 
|- style="background: yellow; text-align: center;"

Revision as of 07:27, 30 July 2010

Introduction

Goal

Goal: To identify artist from music audio (in PCM format).

Datasets

Dataset: Two sets of data were used: Magnatune and USPOP. The audio sampling rates used were either 44.1 KHz or 22.05 KHz (mono). More data information is in the following table.

Dataset Size (@ 44.1 KHz) Number of Training Files Number of Testing Files
Magnatune 35.2 GB 1158 642
USPOP 37.3 GB 1158 653

Results

Overall

OVERALL
Rank Participant Mean of Magnatune Raw Classification Accuracy and USPOP Raw Classification Accuracy
1 Mandel & Ellis 72.45%
2 Bergstra, Casagrande, & Eck (1) 68.57%
3 Bergstra, Casagrande, & Eck (2) 66.71%
4 Pampalk, E. 61.28%
5 West & Lamere 47.24%
6 Tzanetakis, G. 42.05%
7 Logan, B 25.95%


Magnatune Dataset

Magnatune Dataset
Rank Participant Raw Classification Accuracy Normalized Raw classification Accuracy Runtime (s) Machine Confusion Matrix Files
1 Bergstra, Casagrande, & Eck (1) 77.26% 79.64% 24 hours B0 BCE_1_MTeval.txt
2 Mandel & Ellis 76.60% 76.62% 11073 R ME_MTeval.txt
3 Bergstra, Casagrande, & Eck (2) 74.45% 74.51% -- -- BCE_2_MTeval.txt
4 Pampalk, E. 66.36% 66.48% 4272 B1 P_MTeval.txt
5 Tzanetakis, G. 55.45% 55.59% 2632 B0 T_MTeval.txt
6 West & Lamere 53.43% 53.48% 27480 B3 WL_MTeval.txt
7 Logan, B 37.07% 37.10% N/A B3 L_MTeval.txt
8 Lidy & Rauber (SSD+RH) TO * -- -- -- --
8 Lidy & Rauber (RP+SSD) TO * -- -- -- --
8 Lidy & Rauber (RP+SSD+RH) TO * -- -- -- --

USPOP Dataset

USPOP Dataset
Rank Participant Raw Classification Accuracy Normalized Raw classification Accuracy Runtime (s) Machine Confusion Matrix Files
1 Mandel & Ellis 68.30% 67.96% 10240 R ME_USeval.txt
2 Bergstra, Casagrande, & Eck (1) 59.88% 60.90% 24 Hours B0 ME_USeval.txt
3 Bergstra, Casagrande, & Eck (2) 58.96% 58.96% -- -- BCE_2_USeval.txt
4 Pampalk, E. 56.20% 56.03% 4321 B1 P_USeval.txt
5 West & Lamere 41.04% 41.00% 26871 B3 WL_USeval.txt
6 Tzanetakis, G. 28.64% 28.48% 2443 B0 T_USeval.txt
7 Logan, B. 14.83% 14.76% N/A B3 L_USeval.txt
8 Lidy & Rauber (SSD+RH) TO * -- -- -- --
8 Lidy & Rauber (RP+SSD) TO * -- -- -- --
8 Lidy & Rauber (RP+SSD+RH) TO * -- -- -- --


Note: DNC: did not complete ( error in execution). TO: timed out (did not complete within 24 hours).