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
Is there any theoretical explanation why Asymmetrical block better than a (sparse) 3D Convolution kernel? Normally 3D CNN should have more parameters and bigger receptive field
The text was updated successfully, but these errors were encountered:
Is there any theoretical explanation why Asymmetrical block better than a (sparse) 3D Convolution kernel? Normally 3D CNN should have more parameters and bigger receptive field
The text was updated successfully, but these errors were encountered: