Hello,
I am trying to reproduce the results of SOSNet. I have implemented the code with the architecture from github and training on Notredame I get 1.5 % FPR95 on liberty and yosemite instead of ~1%.
So I wanted to ask you some questions about the hyperparameters used.

1. For Adam optimizer, which performed best, did you use a learning rate scheduler or weight decay?
2. How many data did you use to train SOSNet for the Phototour dataset and HPatches?
3. Did you initialize the weights of the model in a certain way?
4. What kind of data augmentation are you performing? Rotation and flipping?

Thank you