Giters

theolepage / sslsv

Framework for training and evaluating self-supervised learning methods for speaker verification.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

a question about labeled data

liyunlongaaa opened this issue · comments

commented

Hi,dear. I read your paper says "Percentage of labeled data"，but I see your code is "the number label of per people". Are they referring to the same, do they only have 100 labels each?

commented

Hi @liyunlongaaa,

Yes, they are referring to the same thing. In the dataset a speaker can have more than 100 labels but in my experiment "the number of labels per speaker" is set to a value between 0 and 100. Thus, 100% is 100 labels, 20% is 20 labels...