feature-aggregation icon indicating copy to clipboard operation
feature-aggregation copied to clipboard
verified publisher icon paschalidoud

Reading a Custom dataset other than olivetti

Open sameerg07 opened this issue 8 years ago • 2 comments

I have tried with OLR dataset which you have written , it works perfectly fine. I wanted to now experiment on other dataset like grimace , i ordered the this grimce same in the way like OLR, but I am not getting any results . Many errors. So can you please write up a python file to do that .

sameerg07 avatar Jul 29 '17 04:07 sameerg07

I am not sure I understand your issue. Are you having problems with the feature aggregation or with reading the dataset?

angeloskath avatar Jul 30 '17 12:07 angeloskath

Yeah , @angeloskath i had doubts on how you have resized the images, in the sense , when i tried doing with the grimace data, it is overfitting and giving very high results. Should we compulsarily graayscale the i images? I wanted to know what exactly the 100 in Vlad(100) mean? like does it vary with the input size and stuff

sameerg07 avatar Jul 30 '17 16:07 sameerg07