- Clone the code (now you should have a folder named MaskTune)
- Inside
Masktune/
createdatasets/
folder - For
catsvsdogs
andinl9 (the Background Challenge)
expriments, insideMaskTune/datasets/
createcatsvsdogs/raw/
andin9l/raw/
folders. For other datasets ignore this step. - Download the dataset you want (you don't need to download cifar10, mnist, and svhn because they will be downloaded automatically).
- For Waterbirds, please download the corrected version of the Waterbirds dataset from here (The original Waterbirds dataset has some label and image noise). Then extract it into the
Masktune/datasets/Waterbirds/
(so inside this folder you should haveimages
folder) - For CelebA, please download
img_align_celeba
folder from here. After extracting it, you should see a folder namedarchive
. Pass this folder's path to the--dataset_dir
in the bash file. - To run an experiment, use the bash files in
MaskTune/bash_files
. First, change the second line of the bash file to the path ofMaskTune
folder (e.g.,downloads/MaskTune
). You have to setbase_dir
to the path ofMaskTune/
folder anddataset_dir
to the path of corresponding dataset (e.g., for celebA set this to{base_dir}/datasets/CelebA/raw
)
sfu-mial / masktune Goto Github PK
View Code? Open in Web Editor NEWCode supporting our NeurIPS paper MaskTune: Mitigating Spurious Correlations by Forcing to Explore