I'm truly honored to have had the opportunity to read your research paper and experiment with the associated code. From the data description in your paper, coupled with my understanding of the code, it seems you might have sliced the Potsdam and Vaihingen datasets into 512x512 image blocks. However, in the code you published on Github, I couldn't find any related data preprocessing sections. Did you conduct any additional processing in this regard? If not, I'm considering using a sliding window approach to slice the Potsdam or Vaihingen data into 512x512 image blocks.
Hello!
I'm truly honored to have had the opportunity to read your research paper and experiment with the associated code. From the data description in your paper, coupled with my understanding of the code, it seems you might have sliced the Potsdam and Vaihingen datasets into 512x512 image blocks. However, in the code you published on Github, I couldn't find any related data preprocessing sections. Did you conduct any additional processing in this regard? If not, I'm considering using a sliding window approach to slice the Potsdam or Vaihingen data into 512x512 image blocks.