DotWang
commented
2 years ago The original Potsdam image is large, so it is needed to be clipped to patches. For preparing the Potsdam dataset, you can refer to https://github.com/open-mmlab/mmsegmentation/blob/master/docs/en/dataset_prepare.md#prepare-datasets. Note there are some differences. We use '3_Ortho_IRRG.zip' and '5_Labels_all.zip'.
For reproduction, we have provided the command lines, config files, training logs, and saved model weights in the readme.md of https://github.com/ViTAE-Transformer/ViTAE-Transformer-Remote-Sensing/tree/main/Semantic%20Segmentation.
As I know, the default mmsegmentation adopts the distributed training. It doesn't affect the final result.
smallsnailrunning
Hello, I am interested in the remote sensing semantic segmentation of this project. I have downloaded all the relevant Potsdam datasets and configured the program running environment, but the reproduction steps of the entire training are still unclear. The downloaded public datasets Whether preprocessing is required, images and label images need to be cut into small pieces. If you don't do distributed training, you can simply remove the distributed settings when training on one card. Do you have a more detailed description of the training steps? The description of the training parameter configuration config file allows us to reproduce the model training. Thank you.