You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have read about Co-learning where we can train model on 3 modalities however at test time we can use only one modality. I am struggling to understand how this will be implemented in a code. Once we train a model with 3 modalities, it will expect 3 modalities at test time. Do we need to handle this scenario by passing zero or some random values for modalities to be dropped. Please help, also any sample implementation of the same.
Thanks a lot for all your awesome repo.
The text was updated successfully, but these errors were encountered:
Hi Paul,
I have read about Co-learning where we can train model on 3 modalities however at test time we can use only one modality. I am struggling to understand how this will be implemented in a code. Once we train a model with 3 modalities, it will expect 3 modalities at test time. Do we need to handle this scenario by passing zero or some random values for modalities to be dropped. Please help, also any sample implementation of the same.
Thanks a lot for all your awesome repo.
The text was updated successfully, but these errors were encountered: