This implementation support for 3 input channels? I got this error when I try:
Traceback (most recent call last):
File "pcanet_plantas.py", line 67, in
ftrain,V,BlkIdx = pcanet.PCANet_train(train_x,PCANet,1)
File "/home/mario/Downloads/PCANet_python-master/PCANet/pcanet.py", line 111, in PCANet_train
V.append(PCA_FilterBank(OutImg, PCANet.PatchSize[stage], PCANet.NumFilters[stage]))
File "/home/mario/Downloads/PCANet_python-master/PCANet/pcanet.py", line 77, in PCA_FilterBank
Rx = Rx + im.dot(im.T)
ValueError: operands could not be broadcast together with shapes (147,147) (49,49)
I supposed that implementation is set to MNIST dataset from the original paper right? But for cifar10, for example, can it works?
This implementation support for 3 input channels? I got this error when I try:
Traceback (most recent call last): File "pcanet_plantas.py", line 67, in
ftrain,V,BlkIdx = pcanet.PCANet_train(train_x,PCANet,1)
File "/home/mario/Downloads/PCANet_python-master/PCANet/pcanet.py", line 111, in PCANet_train
V.append(PCA_FilterBank(OutImg, PCANet.PatchSize[stage], PCANet.NumFilters[stage]))
File "/home/mario/Downloads/PCANet_python-master/PCANet/pcanet.py", line 77, in PCA_FilterBank
Rx = Rx + im.dot(im.T)
ValueError: operands could not be broadcast together with shapes (147,147) (49,49)
I supposed that implementation is set to MNIST dataset from the original paper right? But for cifar10, for example, can it works?