ViTAE-Transformer / SAMRS

The official repo for [NeurIPS'23] "SAMRS: Scaling-up Remote Sensing Segmentation Dataset with Segment Anything Model"
286 stars 14 forks source link

finetune on a custom dataset #8

Open 15926273249 opened 1 year ago

15926273249 commented 1 year ago

I wanna ask how to generate custom datasets using the code or fine-tune the model on custom datasets using the code

DotWang commented 1 year ago

@15926273249 We have provided the instruction, please refer to the readme and do corresponding changes as your case.

15926273249 commented 1 year ago

you mean generate dataset code based on HRSC2016, DOTA, DIOR, FAIR1M , if we need to use other formats of datasets like loveda or Vaihingen Dataset, how should i process to make a correct dataset format to fit the model training

DotWang commented 1 year ago

@15926273249 You mentioned are already segmentation datasets, which can be directly used with the codes of Pretraining and Finetuning.

15926273249 commented 1 year ago

Can you provide the prediction code on the custom datasets of huge images?

DotWang commented 1 year ago

@15926273249 Our code conducts the prediction by slided windows. It can be used for huge images.