Closed kristinbranson closed 1 year ago
My best guess is the issue is that this specific MDN (tensorflow) model was trained on JRC backend which uses a different singularity image from docker. Mayank will update JRC backend to new .sif file. We would need to retrain.
I can track with docker in multiview for a newly JRC trained model.
multianimal branch: Tracking with multiview project errors out using docker backend.