PyTorch Implementation: Code for the paper "Generalizing to Unseen Domains via Adversarial Data Augmentation", NeurIPS 2018. Origin Tensorflow Implementation: https://github.com/ricvolpi/generalize-unseen-domains
Thanks for you sharing beautiful work.
I want to know the function of this link" source /home/users/hlli/anaconda3/bin/activate jigsaw " in run_PACS_alex.sh.
Your code is very comprehensive, but I don't know how to use it. Can you do a tutorial?
Thank you so much!
Hi, can you help me run the code? I couldn't find the datasets for running this version of the method. The following error appears: FileNotFoundError: [Errno 2] No such file or directory: '/home/users/hlli/Datasets/mnist_m/mnistm_with_label.mat'
Hi, your code is very easy to understand, and I really appreciate your contribution. However, I have some questions that I'm not very clear about. When I applied this method to the PACS dataset, I found that the performance was not good, with an accuracy of only about 20%-30% on the target domain. I was wondering if your results were consistent with this? Or do you have any idea about this phenomenon?