Best encoder to get the semantic features

vision
1

Hi,
I have a remote sensing images and I want to get semantic features using an encoder,
My question is what is the best encoder that can I use it to get good features?