MIScnn
MIScnn copied to clipboard
Integrate sample/class weighting
https://github.com/keras-team/keras/issues/3653 https://github.com/keras-team/keras/issues/3653#issuecomment-761085597
https://tensorflow.google.cn/tutorials/images/segmentation?hl=en&skip_cache=true%22
- Implement class dictionary as parameter for data generator
- if parameter != None, compute sample weight according to GitHubUser97 answer (with dynamic n_classes)
- Return sample weight
Add Unittesting for it.
Should also work with patches then.