Deformable-ConvNets-caffe
Deformable-ConvNets-caffe copied to clipboard
unsky
matcaffe error
After I train the network you provide and obtain the Resnet_dcn.caffemodel, I compiled the matlab interface and try to load the caffemodel. However, an error appeared (caffe.layerparameter has no field named deformable-cnvolution parameter). I do not know whether it is caused by the incorrect compilation of matcaffe or the loss of the newly added deformable convolutional layer in caffe_.mexa64. Could you help me?
@haoliyoupai09 you solve the matcaffe error problem?I also want to use deformable conv with matcaffe,looking forward to your reply!!
Yes, I recompiled the matcaffe interface and it worked. Adding a new layaer in caffe deos not influence the use of caffemodel in matlab.
Yes, no error occurs during the compilation of matcaffe and running of code. But the performance is not good with deformable convolutional network. I do not know why.
Yes, no error occurs during the compilation of matcaffe and running of code. But the performance is not good with deformable convolutional network. I do not know why.
@haoliyoupai09 Thanks. What is task you are working on? Are the performance gap significant or slight?
Yes, no error occurs during the compilation of matcaffe and running of code. But the performance is not good with deformable convolutional network. I do not know why.
@haoliyoupai09 I placed one deformable layer at the end of my network. After training, the outputs of this deformable layer are 0. I checked the outputs of the offset layer, which are very large. Do you know why? Could you share more details about how to use this layer? Thanks