search

SarthakYadav / leaf-pytorch

PyTorch implementation of the LEAF audio frontend
64 stars 10 forks source link

Speech Command Dataset #4

Closed wonyangcho closed 2 years ago

wonyangcho commented 2 years ago

Hi @SarthakYadav

For Speech Command Classification, please let me know which dataset you used. I can't find csv file in speech_commands v0.02.tar.gz.

Thansk.

SarthakYadav commented 2 years ago

Yes, there isn't a .csv file in speech_commands v0.02.tar.gz, you have to make it yourself in the following format

files,labels
<filepath>,<string_label>

Another file you'll need is a label map: a .json dictionary that has <string_label>:<class integer> key value pairs.

I use this format for flexibility to not resort to a particular directory structure, as this contains full path of the dataset.

wonyangcho commented 2 years ago

@SarthakYadav Thanks!