[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [MUSIC-IR] test-set for singing voice segmentation

Hi Geoffroy,

You can find the dataset we used for our work on Predominant Melodic Source Separation [1] in the following link:


We also have some of the ground-truth metadata that we can make available (although we may have to organize them a bit: pitch contours extracted from the original and isolated voice tracks using Praat, or the segments where the voice is singing, etc.)



[1] Lagrange, M., Martins, L. G., Murdoch, J., and Tzanetakis, G. (2008). Normalized 
cuts for predominant melodic source separation. IEEE Transactions on Audio, Speech, and Language 
Processing, 16(2). Special Issue on MIR.

Luis Gustavo Martins
Researcher / PhD candidate
Porto, Portugal

On 5 Mar, 2008, at 13:13 , Geoffroy Peeters wrote:

Dear all,

I wonder if there exists any public test set for testing singing voice location algorithms,
or any available annotation of existing test set (such as RWC) annotation into singing voice segments.

Thanks for your help

Geoffroy Peeters
Ircam - R&D
tel: +33/1/
email: peeters@xxxxxxxx

List reminder:
0. ISMIR 2008 will take place September 14-18, 2008 in Philadelphia, PA (USA). The web site of the conference is http://ismir2008.ismir.net/
1. Please do not send HTML documents to the list
2. Please do not send attachments (pictures, Word, etc.) to the list
3. Please do not send commercial ads to the list
4. Reuse of email addresses found on the list for unsolicited mail is forbidden
5. To unsubscribe, mail to "listserv@xxxxxxxx" the following text: unsub music-ir or login to http://lists.ircam.fr/
6. For assistance, mail to "music-ir-request@xxxxxxxxxxxxxxx".
7. The archives of the list are at http://lists.ircam.fr/
8. The collective web sites of the ISMIR conferences and all past proceedings are at http://www.ismir.net/