faster-rcnn-resnet
faster-rcnn-resnet copied to clipboard
Eniac-Xie
ResNet Implementation for Faster-rcnn
who knows why?T-T I'm crazy! I0331 21:30:19.628304 16920 net.cpp:744] Ignoring source layer fc1000 I0331 21:30:19.628332 16920 net.cpp:744] Ignoring source layer prob Solving... Traceback (most recent call last): File "./tools/train_net.py", line...
I noticed that all the parameters of original ResNet are fixed during training. So why do you choose to do so?
Fixed broken download links for all data/scripts files which include faster-rcnn-models, selective-search-data, and imagenet-models.
The fetch_imagenet_model.sh and two other .sh cannot link to the right checksum.So i cannot use the pretrained resnet models,may be because i am in China.someone can help me?thank you very...
Hi, is your open source resnet-fast-rcnn the net result file that you have experimented with the best results?
Please help me! I am having following error when run ` cd $ROOT/lib/ make` `C:\faster-rcnn-resnet\lib>make python setup.py build_ext --inplace Traceback (most recent call last): File "setup.py", line 58, in CUDA...
which caffe you use? Official caffe scale layer do NOT contain batch normalize
I don't find any difference in batchnorm layer between Microsoft's caffe and official caffe. How does the Microsoft's caffe reduces the memory usage? Could you explain with more details? Best!
