I am working on a project related to TCNN and for that I wanted to get an understanding of the execution of your code.
I have used /models/jhmdb/tpn_rec_eval_v3.prototxt network present in the repository and davis_240_320.caffemodel model downloaded from "http://www.cs.ucf.edu/~rhou/files/" and I am getting
F0720 06:04:17.200984 4285 net.cpp:757] Cannot copy param 0 weights from layer 'conv2c'; shape mismatch. Source param shape is 512 128 3 3 3 (1769472); target param shape is 128 128 7 7 (802816). To learn this layer's parameters from scratch rather than copying from a saved net, rename the layer.
Check failure stack trace:
I am working on a project related to TCNN and for that I wanted to get an understanding of the execution of your code.
I have used /models/jhmdb/tpn_rec_eval_v3.prototxt network present in the repository and davis_240_320.caffemodel model downloaded from "http://www.cs.ucf.edu/~rhou/files/" and I am getting
F0720 06:04:17.200984 4285 net.cpp:757] Cannot copy param 0 weights from layer 'conv2c'; shape mismatch. Source param shape is 512 128 3 3 3 (1769472); target param shape is 128 128 7 7 (802816). To learn this layer's parameters from scratch rather than copying from a saved net, rename the layer. Check failure stack trace:
The complete logs are here. https://github.com/gireek/mtcnn_issue/blob/master/issue.txt