I am working with 3D medical data but wanted to use the pre-trained models. So I thought of adding a Conv3D block between the layers of the pre-trained module. And while training the whole model will be trained, the pre-trained weights and the new weights in the 3D Conv blocks.
If someone can guide me on how to do that, I would be really gladful.
I really don’t know if this will work, so if you have some other idea, I will be pleased to hear it, Thanks.