search

PatrickTUM / UnCRtainTS

https://patricktum.github.io/cloud_removal/
47 stars 3 forks source link

Dataset preparation #1

Closed hadilou closed 1 year ago

hadilou commented 1 year ago

Hi, Thanks for the nice work and accompanying article. I got your code running with the SEN12MS-CR-TS dataset but ideally I would like to run inference on my own dataset (let's say from a different region). Do you have a script to prepare the dataset for other locations?

PatrickTUM commented 1 year ago

Hi @hadilou,

great to see you interested in our work! If you'd like to apply our code on any other region of interest, you may try fetching your desired data via Google Earth Engine, or process any other data accordingly. We use Sentinel-1 and Sentinel-2 level 1-C products, which are finally handled as specified in the dataloader. Hoping this helps you! Cheers, Patrick

GKG1312 commented 1 year ago

Hi, Thanks for the nice work and accompanying article. I got your code running with the SEN12MS-CR-TS dataset but ideally, I would like to run inference on my own dataset (let's say from a different region). Do you have a script to prepare the dataset for other locations?

Hi @hadilou , I am trying to run the method for the SEN12MS-CR dataset, but being new to this topic I am a little confused about how to arrange the data as per the test instruction. I want to use the following command line 

python test_reconstruct.py --experiment_name my_first_experiment -root1 path/to/SEN12MSCRtrain --root2 path/to/SEN12MSCRtest --root3 path/to/SEN12MSCR --input_t 3 --region all --export_every 1 --res_dir ./inference --weight_folder ./results

How do I separate data for path/to/SEN12MSCRtrain, path/to/SEN12MSCRtest, and path/to/SEN12MSCR? I have all the data as stored originally with folders and subfolders separating them based on season and sensors.

PatrickTUM commented 1 year ago

Hello @GKG1312

to use this command and the provided dataloader, the referenced data is expected to be in the structure as given by the downloading script. If you wish to perform training or inference on the mono-temporal SEN12MS-CR dataset, then only the flag --root3 is relevant for you, set it to the directory containing your SEN12MS-CR dataset. You can find the meaning of flags briefly explained in the associated parser script. Hoping this helps!

Cheers, Patrick

GKG1312 commented 1 year ago

Thank you @PatrickTUM It was very helpful.

starcksi commented 7 months ago

Hi @hadilou and @GKG1312,

Did you find any way to run inference on new data?