songxinhang / d-cnn Goto Github PK

layer {
name: "fc5"
type: "Convolution"
bottom: "fc4"
top: "fc5"
convolution_param {
num_output: 19
pad: 0
kernel_size: 1
}
}

I'm trying to load the pretrained D-CNN network, and I'm getting the following error

F1101 23:03:41.857909 73 net.cpp:757] Cannot copy param 0 weights from layer 'fc4'; shape mismatch. Source param shape is 512 4096 (2097152); target param shape is 512 256 4 4 (2097152). To learn this layer's parameters from scratch rather than copying from a saved net, rename the layer.

I notice that 512 x 256 x 4 x 4 == 512 x 4096, so it seems that in saving and reloading the network weights the layer was somehow flattened.

To reproduce, I run

import caffe
net = caffe.Net('deploy_D-CNN.prototxt', 'D-CNN.caffemodel', caffe.TEST)

I'm using python 3.5.2 and caffe version 1.0.0

Any ideas how can I counteract this error?