TimesFormer for NTU120? #2072
Unanswered
taliegemen
asked this question in
Q&A
TimesFormer for NTU120?
#2072
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
Hello, I am trying to use dividedST timesformer approach to train NTU120 dataset. The maximum accuracy that I have is around 0.74 top-1 whereseas resnet50slow model achieves around 0.85 performance. The network fully learns the training set (with more than 99 top-1) but overfits on validation set. What to do at this point?
Beta Was this translation helpful? Give feedback.
All reactions