Open bagustris opened 2 years ago
Hi @bagustris.
the assumption of the directory structure is still using EMOVO structure instead of EmoV-DB
I believe it's just the description for EmoV-DB which is incorrect; the script still works with the sorted data. I'll fix this now.
the downloaded dataset was assumed from the mega cloud instead of the original dataset
I'll see whether I can write a script for the original unsorted dataset, but when I last had a look it was somewhat disorganised so will take a more work to process properly.
there is no
veneccoutryinfo.csv
as required in this tool when downloading the dataset from the original source
Now that I think about it, I believe Alan Cowen emailed me this CSV when I emailed him about the data. You might have to email him because I'm not sure whether I'm allowed to make the country and speaker info public.
I tried to follow your experiment with the VENEC dataset, but there is no
veneccoutryinfo.csv
as required in this tool when downloading the dataset from the original source (only wavs and other four .csv files)Also, for EmoV-DB dataset:
Thanks in advance for such a great tool!