export PYTHONPATH=/path/to/Caffe/python:$PYTHONPATH
export PYTHONPATH=/path/to/DLAnalyzer/python:$PYTHONPATH
export PYTHONPATH=/path/to/DLAnalyzer/config:$PYTHONPATH
High-level instructions:
-
Go to the config directory and have a look at the _template.py files and the variables defined in it.
-
To make your own configuration, go to caffe_config.py and selector_config.py, and have a look at the example configurations. This is a good starting point.
-
To run the DLAnalyzer, go to test directory. Here are some key information:
a. csv file (CRF_data.csv in this example) which will include information about the patients (CRF), corresponding images and a set of pictures, which will be used in the algorithm (both training and testing). In the csv file, the corresponding images are used as a .txt file, indicating the path to the iriginal raw images.
b. The actual data can be locating elsewhere
c. The Segnet template model is located at: model_segnet. This is template and all variables in all three files will be set to those indicated in the config file. The template original directory is read only, so a copy with fixed values will be placed inside the working area.
d. The output of the DLAnalyzer will be created in inside the test directory, named as you had indicated in the config file
-
To run the machinery, execute python run.py
export DYLD_LIBRARY_PATH=/usr/local/cuda/lib64:$DYLD_LIBRARY_PATH export PYTHONPATH=~/caffe-segnet-cudnn5/python:$PYTHONPATH export PYTHONPATH=/disk2/nik/Analyzer/python:$PYTHONPATH export PYTHONPATH=/disk2/nik/Analyzer/config:$PYTHONPATH export LD_LIBRARY_PATH="$LD_LIBRARY_PATH:/usr/local/cuda/lib64" export CUDA_HOME=/usr/local/cuda