Baskaran R K R, Link A, Porr B, Franke T (2024)

Soft Matter

DOI of the code: 10.5281/zenodo.8126539

The repository has been archived and is read-only.

• Python 3.10
• Tensorflow & Keras 2.13.0
• OpenCV
• NumPy
• Matplotlib
• tqdm

1. run main.py to train, create, validate and test the model.
2. run plots.py to generate the plots as seen in the paper.

Train, validate and test (native vs chem. mod.) RBCs.

Options:

• FA: Classification of native vs formaldehyde
• DA: Classification of native vs diamide
• GA: Classification of native vs glutaraldehyde
• MIX: Classification of native vs random mix of formaldehyde, diamide, glutaraldehyde

This generates all results in the directory results_<option>.

Runs all option: FA, DA, GA and MIX.

• Foreground: Shows the accuracy and loss.
• Background: nohup ./runall.sh &. You can log out and it will continue.

Loads accuracy_and_loss_values.json and plots accuracy, loss and probability predictions.

Labels the videos, subtracts the background, and returns them as NumPy arrays.

Tests loading videos from the file directory.

Performs background subtraction, displays processed video.