Having troubles in reproducing the experiment on Recurrent Variational Network: A Deep Learning Inverse Problem Solver applied to the task of Accelerated MRI Reconstruction (CVPR 2022)

NKI-AI / direct

Deep learning framework for MRI reconstruction

https://docs.aiforoncology.nl/direct

Apache License 2.0

236 stars 41 forks source link

Having troubles in reproducing the experiment on Recurrent Variational Network: A Deep Learning Inverse Problem Solver applied to the task of Accelerated MRI Reconstruction (CVPR 2022) #201

Closed Winnie-xwj closed 2 years ago

Winnie-xwj commented 2 years ago

I have followed the instructions on the https://docs.aiforoncology.nl/direct/colab.html and successfully installed the 'direct'. Then, according to project folder https://docs.aiforoncology.nl/direct/cvpr2022_recurrentvarnet.html, I pasted the command to Colab, but there is still something wrong.

georgeyiasemis commented 2 years ago

Hi @Winnie-xwj, the command you are running seems wrong. The correct usage is as follows:


direct train <PATH_TO_TRAINING_DATA_DIRECTORY>  <PATH_TO_VAL_DATA_DIRECTORY> <experiment_directory> --num-gpus <number_of_gpus> --cfg <path_or_url_to_yaml_file> [--other-flags]

georgeyiasemis commented 2 years ago

To dowload the Multi_Channel Calgary Campinas Dataset check out: https://portal.conp.ca/dataset?id=projects/calgary-campinas
To see how to run experiments using DIRECT check out: https://docs.aiforoncology.nl/direct/training.html
To reproduce experiments for the RecurrentVarNet paper check out https://docs.aiforoncology.nl/direct/cvpr2022_recurrentvarnet.html