search

PRBonn / bonnet

Bonnet: An Open-Source Training and Deployment Framework for Semantic Segmentation in Robotics.
GNU General Public License v3.0
323 stars 89 forks source link

validation matrix #30

Closed cshort101 closed 6 years ago

cshort101 commented 6 years ago

Hi, I'm wondering if there's any confusion matrix output from either training or validation to see the per class accuracy, IoU, etc?

tano297 commented 6 years ago

Hi, sorry for the delay, I've been out and about.

Using the cnn_dataset_performance.py script you can get the classwise metrics of an already trained model. The confusion matrix is not published, but it is calculated, so you should be able to print if by making the output of the IoU calculation a bit more verbose, if you need this I can tell you where to look.

cshort101 commented 6 years ago

Thanks for the reply! I guess the confusion matrix is something calculated as 2D histogram in abstract_net.py?