Home

Steps to process images with CDeep3M2:

• Training: train the CNN to recognize objects in 3D image stacks (or 2D images)

• Prediction: image segmentation

• Transfer Learning: adapt a pre-trained model to a new dataset or different object

For prediction and transfer learning you can use either a pre-trained model from the modelzoo or your own model generated during training.