active-segmentation
active-segmentation copied to clipboard
HealthML
ActiveSegmentation: A Simulation Framework for Benchmarking Active Learning Strategies for 3D Medical Image Segmentation
When I use Multi-GPU for Training, it has error:in dump ForkingPickler(file, protocol).dump(obj) AttributeError: Can't pickle local object 'TorchHistory.add_log_parameters_hook..'
https://github.com/HealthML/active-segmentation/blob/main/src/datasets/doubly_shuffled_nifti_dataset.py#L499 this line will throw an error when self.dim is not 2.
- [ ] Make sure all dependencies are in the conda env - [ ] Test if the stuff in the readme actually works
Could be implemented in pytorch. Testing will be tedious. Problems: They compare target and predicted surface if the distance between target predicted surface is above a certain threshold. This threshold...
Query strategies are currently not working for 3D models. **Random**: - Samples a random scan without a slice, which results in labeling slice `0` of that scan - Could be...
