Difference between revisions of "2005:Audio Key Finding Results"

From MIREX Wiki
(→‎Results)
 
(4 intermediate revisions by the same user not shown)
Line 1: Line 1:
 +
==Introduction==
 +
 +
===Goal===
 +
The evaluation of key finding algorithms applied to audio sound files
 +
 +
===Dataset===
 +
1,252 audio files synthesized from MIDI Note: There is a close relationship (same musical datasets) between this contest and the Symbolic Key Finding contest. Here is a link to the Symbolic Key Finding results.
 +
 +
Two databases used: Winamp synthesized audio (w) and Timidity with Fusion soundfonts synthesized audio (t). Each database is approximately 3.1 gigabytes for a total of 6.2 gigabytes of audio files.
 +
 +
The composite score is calculated by averaging the Winamp and Timidity scores.
 +
 +
==Results==
 
{| border="1" cellspacing="0"
 
{| border="1" cellspacing="0"
 
|- style="background: yellow;"
 
|- style="background: yellow;"
Line 36: Line 49:
 
|----
 
|----
 
|1
 
|1
−
|?zmirli, Ö.
+
|[https://www.music-ir.org/mirex/abstracts/2005/izmirli.pdf Izmirli, Ö. ]
−
|89.55%
+
|align="right " | 89.55%
−
|1188.8
+
|align="right " | 1188.8
−
|1122.9
+
|align="right " | 1122.9
−
|89.4%
+
|align="right " | 89.4%
−
|89.7%
+
|align="right " | 89.7%
−
|1086
+
|align="right " | 1086
−
|1089
+
|align="right " | 1089
−
|36
+
|align="right " | 36
−
|42
+
|align="right " | 42
−
|38
+
|align="right " | 38
−
|31
+
|align="right " | 31
−
|17
+
|align="right " | 17
−
|18
+
|align="right " | 18
−
|75
+
|align="right " | 75
−
|72
+
|align="right " | 72
−
|15284
+
|align="right " | 15284
−
|16354
+
|align="right " | 16354
−
|Y
+
|align="right " | Y
 
|----
 
|----
 
|2
 
|2
−
|Purwins & Blankertz
+
|[https://www.music-ir.org/mirex/abstracts/2005/purwins.pdf Purwins & Blankertz ]
−
|89.00%
+
|align="right " | 89.00%
−
|1122.4
+
|align="right " | 1122.4
−
|1106.5
+
|align="right " | 1106.5
−
|89.6%
+
|align="right " | 89.6%
−
|88.4%
+
|align="right " | 88.4%
−
|1090
+
|align="right " | 1090
−
|1060
+
|align="right " | 1060
−
|44
+
|align="right " | 44
−
|72
+
|align="right " | 72
−
|24
+
|align="right " | 24
−
|21
+
|align="right " | 21
−
|16
+
|align="right " | 16
−
|21
+
|align="right " | 21
−
|78
+
|align="right " | 78
−
|78
+
|align="right " | 78
−
|45003
+
|align="right " | 45003
−
|44232
+
|align="right " | 44232
−
|R
+
|align="right " | R
 
|----
 
|----
 
|3
 
|3
−
|Gómez, E. (start)
+
|[https://www.music-ir.org/mirex/abstracts/2005/gomez.pdf Gómez, E. (start)]
−
|86.05%
+
|align="right " | 86.05%
−
|1081.9
+
|align="right " | 1081.9
−
|1072.9
+
|align="right " | 1072.9
−
|86.4%
+
|align="right " | 86.4%
−
|85.7%
+
|align="right " | 85.7%
−
|1048
+
|align="right " | 1048
−
|1034
+
|align="right " | 1034
−
|35
+
|align="right " | 35
−
|44
+
|align="right " | 44
−
|38
+
|align="right " | 38
−
|43
+
|align="right " | 43
−
|25
+
|align="right " | 25
−
|20
+
|align="right " | 20
−
|106
+
|align="right " | 106
−
|111
+
|align="right " | 111
−
|1560
+
|align="right " | 1560
−
|1531
+
|align="right " | 1531
−
|B 0
+
|align="right " | B 0
 
|----
 
|----
 
|4
 
|4
−
|Gómez, E. (global)
+
|[https://www.music-ir.org/mirex/abstracts/2005/gomez.pdf Gómez, E. (global)]
−
|85.90%
+
|align="right " | 85.90%
−
|1076.1
+
|align="right " | 1076.1
−
|1073.8
+
|align="right " | 1073.8
−
|86.0%
+
|align="right " | 86.0%
−
|85.8%
+
|align="right " | 85.8%
−
|1019
+
|align="right " | 1019
−
|1015
+
|align="right " | 1015
−
|69
+
|align="right " | 69
−
|73
+
|align="right " | 73
−
|62
+
|align="right " | 62
−
|59
+
|align="right " | 59
−
|20
+
|align="right " | 20
−
|23
+
|align="right " | 23
−
|82
+
|align="right " | 82
−
|82
+
|align="right " | 82
−
|2041
+
|align="right " | 2041
−
|1971
+
|align="right " | 1971
−
|B 0
+
|align="right " | B 0
 
|----
 
|----
 
|5
 
|5
−
|Pauws, S.
+
|[https://www.music-ir.org/mirex/abstracts/2005/pauws.pdf Pauws, S. ]
−
|85.00%
+
|align="right " | 85.00%
−
|1055.1
+
|align="right " | 1055.1
−
|1072.8
+
|align="right " | 1072.8
−
|84.3%
+
|align="right " | 84.3%
−
|85.7%
+
|align="right " | 85.7%
−
|1019
+
|align="right " | 1019
−
|1034
+
|align="right " | 1034
−
|20
+
|align="right " | 20
−
|23
+
|align="right " | 23
−
|67
+
|align="right " | 67
−
|69
+
|align="right " | 69
−
|30
+
|align="right " | 30
−
|33
+
|align="right " | 33
−
|116
+
|align="right " | 116
−
|93
+
|align="right " | 93
−
|503
+
|align="right " | 503
−
|507
+
|align="right " | 507
−
|G
+
|align="right " | G
 
|----
 
|----
 
|6
 
|6
−
|Zhu, Y.
+
|[https://www.music-ir.org/mirex/abstracts/2005/zhu.pdf Zhu, Y.]
−
|83.25%
+
|align="right " | 83.25%
−
|1066.2
+
|align="right " | 1066.2
−
|1017.7
+
|align="right " | 1017.7
−
|85.2%
+
|align="right " | 85.2%
−
|81.3%
+
|align="right " | 81.3%
−
|1034
+
|align="right " | 1034
−
|964
+
|align="right " | 964
−
|38
+
|align="right " | 38
−
|66
+
|align="right " | 66
−
|28
+
|align="right " | 28
−
|47
+
|align="right " | 47
−
|24
+
|align="right " | 24
−
|33
+
|align="right " | 33
−
|128
+
|align="right " | 128
−
|142
+
|align="right " | 142
−
|25233
+
|align="right " | 25233
−
|24039
+
|align="right " | 24039
−
|Y
+
|align="right " | Y
 
|----
 
|----
 
|7
 
|7
−
|Chuan & Chew
+
|[https://www.music-ir.org/mirex/abstracts/2005/chuan.pdf Chuan & Chew]
−
|79.10%
+
|align="right " | 79.10%
−
|1002.3
+
|align="right " | 1002.3
−
|977.3
+
|align="right " | 977.3
−
|80.1%
+
|align="right " | 80.1%
−
|78.1%
+
|align="right " | 78.1%
−
|937
+
|align="right " | 937
−
|905
+
|align="right " | 905
−
|83
+
|align="right " | 83
−
|95
+
|align="right " | 95
−
|66
+
|align="right " | 66
−
|68
+
|align="right " | 68
−
|20
+
|align="right " | 20
−
|22
+
|align="right " | 22
−
|146
+
|align="right " | 146
−
|162
+
|align="right " | 162
−
|3299
+
|align="right " | 3299
−
|3468
+
|align="right " | 3468
−
|R
+
|align="right " | R
 
|----
 
|----
 
|}
 
|}

Latest revision as of 16:34, 2 August 2010

Introduction

Goal

The evaluation of key finding algorithms applied to audio sound files

Dataset

1,252 audio files synthesized from MIDI Note: There is a close relationship (same musical datasets) between this contest and the Symbolic Key Finding contest. Here is a link to the Symbolic Key Finding results.

Two databases used: Winamp synthesized audio (w) and Timidity with Fusion soundfonts synthesized audio (t). Each database is approximately 3.1 gigabytes for a total of 6.2 gigabytes of audio files.

The composite score is calculated by averaging the Winamp and Timidity scores.

Results

Rank Participant Composite Percentage Score Total Score Percentage Score Correct Keys Perfect 5th Errors Relative Major/Minor Errors Parallel Major/Minor Errors Other Errors Runtime (s) Machine
w t w t w t w t w t w t w t w t
1 Izmirli, Ö. 89.55% 1188.8 1122.9 89.4% 89.7% 1086 1089 36 42 38 31 17 18 75 72 15284 16354 Y
2 Purwins & Blankertz 89.00% 1122.4 1106.5 89.6% 88.4% 1090 1060 44 72 24 21 16 21 78 78 45003 44232 R
3 Gómez, E. (start) 86.05% 1081.9 1072.9 86.4% 85.7% 1048 1034 35 44 38 43 25 20 106 111 1560 1531 B 0
4 Gómez, E. (global) 85.90% 1076.1 1073.8 86.0% 85.8% 1019 1015 69 73 62 59 20 23 82 82 2041 1971 B 0
5 Pauws, S. 85.00% 1055.1 1072.8 84.3% 85.7% 1019 1034 20 23 67 69 30 33 116 93 503 507 G
6 Zhu, Y. 83.25% 1066.2 1017.7 85.2% 81.3% 1034 964 38 66 28 47 24 33 128 142 25233 24039 Y
7 Chuan & Chew 79.10% 1002.3 977.3 80.1% 78.1% 937 905 83 95 66 68 20 22 146 162 3299 3468 R
Retrieved from "https://music-ir.org/mirex/w/index.php?title=2005:Audio_Key_Finding_Results&oldid=7622"