IshitaTakeshi/PCANet

only 41% accuracy for our own dataset (4 classes-each102 images)

arunadevikaruppasamy opened this issue · 1 comments

I have used this PCANet code for the dataset which contains cat(102),dog(102),horses(102),human images(102) of total 808 images.
For training , i gave 600 images and 208 images for testing. but showed poor performance than tradition CNN network and VGG16
please can you help me to resolve this issue

Not an issue of this software