-
Notifications
You must be signed in to change notification settings - Fork 5
/
Copy pathMakefile
52 lines (28 loc) · 2.03 KB
/
Makefile
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
############### Dataset processing commands ##################
createXrayDataset:
python dataset_tool.py create_from_images datasets/xray_frontal_train /research/boundaryCrossing/chestxray_raz/square_frontal_train
#last argument contains a folder with all pngs in the same power-of-2 resolution.
createBrainDataset:
python dataset_tool.py create_from_images datasets/brains_test_mono /razspace/brain_datasets/pngs_test
######### Model training commands #################
trainBrains:
python run_training.py --num-gpus=4 --data-dir=datasets --config=config-e --dataset=brains_train_mono --mirror-augment=true
######### Image generation (post-training) #################
genFFHQ:
python run_generator.py generate-images --seeds=0-200 --truncation-psi=0.5 --network=results/00344-stylegan2-ffhq_train-4gpu-config-e/network-snapshot-007127.pkl
genBrains:
python run_generator.py generate-images --seeds=0-200 --truncation-psi=0.5 --network=results/00296-stylegan2-brains_train_mono-4gpu-config-e/network-snapshot-001451.pkl
genXray:
python run_generator.py generate-images --seeds=0-200 --truncation-psi=0.5 --network=results/00341-stylegan2-xray_frontal_train-4gpu-config-e/network-snapshot-007946.pkl
project:
python run_projector.py project-real-images --dataset=xray_ood --data-dir=datasets --network=results/00037-stylegan2-xray_1k_v2-8gpu-config-e/network-snapshot-005488.pkl
######### Image reconstruction #################
# recontype = "super-resolution", "inpaint"
super-resolutionFFHQ:
python recon.py recon-real-images --input=datasets/ffhq --tag=ffhq --network=dropbox:ffhq.pkl --recontype=super-resolution
# this toy dataset contains chst x-ray images from wikipedia
inpaintXray:
python recon.py recon-real-images --input=datasets/xray --masks=masks/1024x1024 --tag=xray --network=dropbox:xray.pkl --recontype=inpaint
# brains contains test images from the OASIS dataset only
inpaintBrains:
python recon.py recon-real-images --input=datasets/brains --masks=masks/256x256 --tag=brains --network=dropbox:brains.pkl --recontype=inpaint