Closed edgark31 closed 1 month ago
Created a function in Utils to split train val and test sets so that both YOLO and RetinaNet models split data the same way
A long time ago, we had functions that did this in AxonDeepSeg. Maybe there might be some useful insights in those old files too:
https://github.com/axondeepseg/axondeepseg/blob/v3.3.0/notebooks/01-guide_dataset_building.ipynb
https://github.com/axondeepseg/axondeepseg/tree/v3.3.0/AxonDeepSeg/data_management
And the tests,
https://github.com/axondeepseg/axondeepseg/blob/v3.3.0/test/data_management/test_dataset_building.py
Created a function in Utils to split train val and test sets so that both YOLO and RetinaNet models split data the same way
A long time ago, we had functions that did this in AxonDeepSeg. Maybe there might be some useful insights in those old files too:
https://github.com/axondeepseg/axondeepseg/blob/v3.3.0/notebooks/01-guide_dataset_building.ipynb
https://github.com/axondeepseg/axondeepseg/tree/v3.3.0/AxonDeepSeg/data_management
And the tests,
https://github.com/axondeepseg/axondeepseg/blob/v3.3.0/test/data_management/test_dataset_building.py
Thank you! I'll take a look at them and talk to @hermancollin about it and the other changes