RPN_BF的配置

1、 download caffe-RPN_BF code from https://github.com/zhangliliang/caffe/tree/RPN_BF

2、 为了使faster-RCNN支持cudnn V5:
(1) 用caffe-dir/include/caffe/util/cudnn.hpp文件替换 caffe-RPN_BF中的cudnn.hpp;
(2) 用caffe-dir/src/caffe/layers/cudnn_*文件(所有以cudnn开头的文件)替换caffe-RPN_BF中的相应文件。

3、 编译caffe-RPN_BF:

cp Makefile.config.example Makefile.config
# Adjust Makefile.config (for example, if using Anaconda Python, or if cuDNN is desired)
make all
make test
make runtest
make matcaffe

(参考:https://github.com/zhangliliang/caffe/tree/RPN_BF)

4、
download RPN_BF-RPN-pedestrian code from https://github.com/zhangliliang/RPN_BF/tree/RPN-pedestrian

5、 编译运行RPN_BF-RPN-pedestrian:
(1) Copy Caffe_DIR/matlab/+caffe/private/caffe_.mexa64 to RPN_BF_DIR/external/caffe/matlab/caffe_faster_rcnn/.
(2) 按照https://github.com/zhangliliang/RPN_BF/tree/RPN-pedestrian步骤执行test 步骤出现错误:MATLAB/R2014a/bin/glnxa64/../../sys/os/glnxa64/libstdc++.so.6: version GLIBCXX_3.4.21
错误原因:matlab 所引用的库与caffe不同(具体参考博客:http://blog.csdn.net/fangbinwei93/article/details/52865461)
解决方法:

sudo rm /usr/local/MATLAB/R2016b/sys/os/glnxa64/libstdc++.so.6
sudo ln -s /usr/lib/x86_64-linux-gnu/libstdc++.so.6 /usr/local/Matlab/R2013a/sys/os/glnxa64/libstdc++.so.6 

本文参考博客:
http://blog.csdn.net/fangbinwei93/article/details/52865461
http://www.cnblogs.com/LiuSY/p/6782560.html

你可能感兴趣的:(深度学习)