What is test/train split from the given 3799 samples? #8
Description
How are the train and test set split from 3799 examples? What are the indices for train/test split?
Copied from page 7: D Benchmark Dataset Statistics and Shape
In total, 3799 annotated patches are provided without considering any overlap. We use 959 patches to train selected state-of-the-art DL segmentation models, while the remaining 2840 patches are used for evaluation in the experiments.