ruizengalways / fod-net Goto Github PK

hi, thank you for sharing your work, but I'm really interested in how your training phase looks like, in your paper you mentioned that your batch size is 64, what I want to know is, a batch here means 64 FODs in a batch or 64 patches in a batch?

If it is 64 patches in a batch, I noticed that your learning rate is 0.01, which is, in my point of view, to large, and make the network learning painfully. If the batch is 64 FODs, how did you solve the memory issues.

Looking forward for your reply.

Kind regards
Jia

I found your paper "FOD-Net:A deep learning method for fiber orientation distribution angular super resolution ", I feel very inspired, and I found that you have generously shared your source code on github,However, I got poor results in the training and testing process. I think there is something wrong with my data processing or program setting. I hope to get your training model and data case.The links in github to the pretrained weights and dataset have expired. Are you able to provide new links?

The links to the pretrained weights and dataset have expired. Are you able to provide new links?

If you're looking for a place to host datasets/weights for research papers, where the links don't expire, then I would recommend zenodo.org.