Datasets:

projecte-aina
/

festcat_trimmed_denoised

Tasks:

Modalities:

Formats:

Languages:

Size:

ArXiv:

DOI:

Libraries:

License:

Dataset card Data Studio Files Files and versions Community

AlexK-PL commited on May 17, 2024

Commit

a9f91fe

·

verified ·

1 Parent(s): b29f692

Update README.md

Files changed (1) hide show

README.md +1 -1

README.md CHANGED Viewed

@@ -58,7 +58,7 @@ We processed the data of the Catalan Festcat with the following recipe:
   - [CleanUNet](https://github.com/NVIDIA/CleanUNet) - [arXiv](https://arxiv.org/abs/2202.07790)
 We kept the same number of samples, filenames, also the original anonymized speaker IDs and transcriptions.
-Our dataset version, after trimming, accumulates a total of 22.91h divided by speaker ID as follows:
 - bet (female): 0.97h
 - mar (female): 0.96h

   - [CleanUNet](https://github.com/NVIDIA/CleanUNet) - [arXiv](https://arxiv.org/abs/2202.07790)
 We kept the same number of samples, filenames, also the original anonymized speaker IDs and transcriptions.
+Our dataset version, after trimming, accumulates a total of 22.91h (divided by speaker IDs) as follows:
 - bet (female): 0.97h
 - mar (female): 0.96h