2017:Drum Transcription - Revision history

Richard Vogl: /* Submission closing date */

2017-09-01T08:40:24Z

‎Submission closing date

Richard Vogl: /* Training Data */

2017-09-01T08:35:22Z

‎Training Data

Richard Vogl: /* Training Data */

2017-09-01T08:34:51Z

‎Training Data

Richard Vogl: /* Data */

2017-09-01T08:34:03Z

‎Data

Richard Vogl: /* Data */

2017-09-01T08:33:16Z

‎Data

Richard Vogl: /* Data */

2017-08-30T09:06:27Z

‎Data

Richard Vogl: added train data link

2017-08-14T13:34:22Z

added train data link

Richard Vogl at 12:21, 31 July 2017

2017-07-31T12:21:30Z

Richard Vogl: No additional traning data as discussed with Carl and Chih-Wei

2017-06-19T11:41:07Z

No additional traning data as discussed with Carl and Chih-Wei

Richard Vogl at 14:59, 8 June 2017

2017-06-08T14:59:03Z

← Older revision		Revision as of 08:40, 1 September 2017
Line 166:		Line 166:
	==Submission closing date==		==Submission closing date==

−	'''September 4th 2017'''	+	<strike>'''September 4th 2017'''</strike>
		+	'''September 11th 2017'''

@@ Line 42: / Line 42: @@
 * A representative random subset of the data will be made available to all participants in advance of the evaluation - please contact the task captains!
-* This data can be used by the participants as they please.
+* Training data can be used by the participants as they please.
-* This data will not be used again during the evaluation
+* Training data will not be used again during the evaluation.
 * Usage of additional training data is discouraged. If additional training data is used, please note so in the submission.

@@ Line 41: / Line 41: @@
 ===Training Data===
-* A representative random subset of the data will be made available to all participants in advance of the evaluation [https://drive.google.com/open?id=0B5QVjGCSDYW1VkZwUEhRSWR5VVk here]
+* A representative random subset of the data will be made available to all participants in advance of the evaluation - please contact the task captains!
 * This data can be used by the participants as they please.
 * This data will not be used again during the evaluation

@@ Line 20: / Line 20: @@
 Additionally three new datasets will be used. They contain polyphonic music of different genres, as well as drum only tracks, and some tracks without drums:
-* RBMA set (35 full length, polyphonic tracks, electronically produced and recorded)
+* RBMA set (35 full length, polyphonic tracks, electronically produced and recorded, manually annotated and double-checked)
-* MEDLEY set (23 full length tracks, recorded)
+* MEDLEY set (23 full length tracks, recorded, manually annotated and double-checked)
 * GEN set (synthesized MIDI drum tracks and loops without accompaniment)

@@ Line 11: / Line 11: @@
 ==Data==
-For evaluation 6 datasets will be used.
+For evaluation 6 (5) different datasets will be used.
 By the time the evaluation is run, we hope to have the three datasets from the 2005 drum detection MIREX task as a baseline.
 Currently we only have the set provided by Christian Dittmar and Koen Tanghe.

@@ Line 13: / Line 13: @@
 For evaluation 6 datasets will be used.
 By the time the evaluation is run, we hope to have the three datasets from the 2005 drum detection MIREX task as a baseline.
-Currently we only have the set provided by Christian Dittmar.
+Currently we only have the set provided by Christian Dittmar and Koen Tanghe.
 * CD set
-* (KT set)
+* KT set
 * (GM set)
 Additionally three new datasets will be used. They contain polyphonic music of different genres, as well as drum only tracks, and some tracks without drums:
-* RV set (35 full length, polyphonic tracks, electronically produced and recorded)
+* RBMA set (35 full length, polyphonic tracks, electronically produced and recorded)
-* CW-CS set (23 full length tracks, recorded [TODO])
+* MEDLEY set (23 full length tracks, recorded)
-* MM set (20 synthesized MIDI drum tracks)
+* GEN set (synthesized MIDI drum tracks and loops without accompaniment)

@@ Line 164: / Line 164: @@
 Software: Preferred Python. May be Matlab, Sonic Annotator.
 ==Submission closing date==
-[TODO]

@@ Line 2: / Line 2: @@
 Drum transcription is the task of detecting the positions in time and labeling the drum class of drum instrument onsets in polyphonic music.
-The information gained is very useful for several applications and can also be useful for further high-level MIR tasks.
+This information is a prerequisite for several applications and can also be used for other high-level MIR tasks.
-Since in the recent past, several new approaches for this task have been published we propose to reboot this task.
+Due to several new approaches recently being presented we propose to reintroduce this task.
 We will mainly stick to the mode used in the first edition in 2005, but new datasets will be used.
 Only the three main drum instruments of drum kits for western pop music are considered.
@@ Line 42: / Line 42: @@
 * A representative random subset of the data will be made available to all participants in advance of the evaluation
-* This data can be used by the participants as they please, additional training data may be used [TODO]
+* This data can be used by the participants as they please.
 * This data will not be used again during the evaluation
 ==I/O format==
@@ Line 132: / Line 133: @@
 Conditions:
 * The actual drum sounds (sound samples) used in any of the input audio are not public and not used for training.
-* Participants who provided data and who need in-advance training or tuning, should only use the data made available to all participants by the organizers - and optionally additional other data.
+* Participants are encouraged to only use the provided training data for training and parameter optimization.
-If this is not possible, they should explicitly state that they used their own data that was donated to be used in the MIREX evaluation so that this is known in public, and that they can be put in a separate category. In this case it would be favorable to submit two versions: one trained with the public data only, and one trained using all of their own data. The point is that this must be clear to everyone so that this is known for interpreting the evaluation results correctly.
+If this is not possible, it should explicitly be stated that and which additional data was used. In this case it would be favorable to submit two versions: one trained with the public data only, and one trained using additional data.