nkkk98 / CycleGAN-medical-image-segmentation

Implementation of CycleGAN for unsupervised image segmentaion, performed on brain tumor scans

CycleGAN

As proposed by Jun-Yan Zhu in Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks.

Prerequisites

• tensorflow r1.1
• numpy 1.11.0
• scipy 0.17.0
• pillow 3.3.0

Train

python main.py --dataset_dir=med_image

Test

python main.py --dataset_dir=med-image --phase=test --which_direction=AtoB

References

• CycleGAN implementation, https://github.com/XHUJOY/CycleGAN-tensorflow
• Torch CycleGAN, https://github.com/junyanz/CycleGAN