Data and accuracy

[dpny518]

Hi, do you have some background on the data source, how much data was used to train, is this data available, and the accuracy

thank you

[riebling]

I didn't develop this code, only made it into a GitHub repository, written by Yun Wang at CMU's Language Technologies Institute. I can say that the data was trained on the open source child speech data from http://homebank.talkbank.org, especially the VanDam data. It would be a good thing to find and add this information. Also, this repository is no longer being maintained, the one at https://github.com/srvk/Yunitator is likely to be more current & receive updates

[riebling]

Follow this Issue: srvk/DiViMe#62