Closed amartia closed 7 years ago
I'm trying to use the bilinear2D function in layers/ourlayers.py and I'm having an error because it wants to use theano but it is not always imported
dim_ordering = K.image_dim_ordering() if dim_ordering == 'th': import theano from theano import tensor as T
Is there a way to use this function with the Tensorflow backend or do we have to implement it?
For the moment, I've replaced it by the UpSampling2D layer but I would like to use the bilinear one.
In any case, what is the proper way to call this function?
Thank you
I have uploaded a bilinear initializer in intializations/initializations.py Forget about the previous implementation that only worked for Theano.
I'm trying to use the bilinear2D function in layers/ourlayers.py and I'm having an error because it wants to use theano but it is not always imported
Is there a way to use this function with the Tensorflow backend or do we have to implement it?
For the moment, I've replaced it by the UpSampling2D layer but I would like to use the bilinear one.
In any case, what is the proper way to call this function?
Thank you