/TPRNet

code for "Fusing deep and handcrafted features for intelligent recognition of uptake patterns on thyroid scintigraphy".

Primary LanguagePythonMIT LicenseMIT

TPRNet

code for "Fusing deep and handcrafted features for intelligent recognition of uptake patterns on thyroid scintigraphy".

Requirements

  • PyTorch 1.0+
  • Python 3.x

Running

train:

python train.py --config cfgs/tprnet.yaml --gpu 0 --net tprnet --model tpr_model

eval:

python eval.py --config cfgs/tprnet.yaml --gpu 0 --net tprnet --model tpr_model

Datasets

The data folder contains a portion of our dataset.