Infectio-photorealistic

Photorealistic output for Infectio. Code to create training dataset and training models to transfer the simplistic x,y positions output of simulation runs in Infectio to realistic looking videos of virus spread in plaque experiements.

Demo

Shows result for model trained for around one milion steps. Left: GT, Middle: model output, right: GT - ouput

How to use

Dataset

Training

Set train parameters in config.py, then run python train.py.

on Slurm (hemera)

Prediction

To Do

References

reference for pix2pix paper Image-to-Image Translation with Conditional Adversarial Networks