abner2015

I had the same error and my tensorflow version is 1.2.1

http://www.robots.ox.ac.uk/~luca/cfnet.html

http://pan.baidu.com/s/1hqGbyQS

http://pan.baidu.com/s/1hqGbyQS

you can change in the file of VOCinit.m . the dir is VOCdevkit\VOCcode

把它注释掉吧

tf2caffe.py 里面，网络名称对应就可以

我这个是做模型转换的，首先是需要有训练好的模型，也就是网络结构是定义好的，所以网络可以根据自己需要来定义，定义完训练。如果是有模型文件，那不能随便定义

The first get the network parameter of tf model , and then get data of tf model . Finally, assign the data to the layer corresponding of caffemodel.

Do as u say, error does not occur. But, other error is happend. ./src/mxnet_my_c_predict_api.cpp:11:28: fatal error: mxnet/symbolic.h: no file or directory #include ^ compilation terminated. make: *** [objs/mxnet_my_c_predict_api.o] error 1