Repository for the code used in " A Deep Learning Approach for Semantic Segmentation of Unbalanced Data in Electron Microscopy of Catalytic Materials."
Python scripts for :
- U-Net model
- Train model
- Evaluate model: DSC, recall, precision scores
Training and validation data sets include ground-truth images and corresponding annotations.
-
Number of classes : 3
Class labels :
Background/Pores : 0, γ-Alumina : 1, Pt nanoparticles : 2
-
Image sets are 512x512 pixels patches.
45 images for training and 15 images for validation
Published paper can be found at: https://www.nature.com/articles/s41598-022-16429-3