it's for drawing&calculate P-R curve/LOSS curve/AP/mAP and store data as .mat for further use
please clone the reposity in your model folders
# Make sure to clone with --recursive
git clone --recursive https://github.com/rbgirshick/py-faster-rcnn.git
this is a really small part code for training loss curve especially in yolo
To rercord the log during training your model, please run the code
cd $darknet(your model file)
script -R log.txt
Once your traning has been done, please don't forget to stop the log by ^c
or exit()
Open your matlab with sudo matlab
and change the train_log_file
(it don't need to change usually)
It's important to change the code [~, string_output] = dos(['cat ', train_log_file, ' | grep "avg," | awk ''{print $3}'''])
so that it can read your own model (if you are using yolo, please don't change it). This is used for extracting the loss
output with dos
command. But in the first version, the genaration is not extracted.
After normalized the code, this script can be run
to plot