Nils Lehmann

Inconsistencies between handling of bands in datasets

10

I am currently working on adding plotting methods based on #253. For datasets that contain Sentinel multispectral imagery, I have two questions. 1. In the CV4AKenyaCropType, the band names can...

Radiant MLHub Crop Type Datasets

5

This PR "superseeds" #511, because I found that Radiant ML Hub has multiple crop type datasets that follow almost the exact same format. Only the label.geojson files differ and hold...

Add STACAPI dataset

31

This PR closes #403 by adding a STACAPI dataset. I will try to explain my thinking for this proposal: Here are some explanations for choices I made: - constructor takes...

Chesapeake CVPR only download state data

3

This PR closes #452 by changing Chesapeake CVPR to work with a selection of states only. TODO: - [ ] add urls, zipfilenames, md5s for new download structure - [...

Retrieve Sentinel Imagery as corresponding input for Geospatial datasets

5

One thing that I find not so straight forward with TorchGeo is finding/using input imagery for the various Geospatial datasets that only provide labels (this could also just be me,...

Functorch Usecase Jacobian computation

10

Hi, Thank you for this exciting work. I will try to explain a use case that I hope will be possible with functorch because the naive way is just extremely...

Model Checkpointing

1

**Describe the bug** In [main.py](https://github.com/weecology/DeepForest/blob/02aac1fd8105f18e63c321be170c2c55d82a51c6/deepforest/main.py#L127), the `enable_checkpointing=False` is hard coded. When I pass a lightning ModelCheckpoint to `create_trainer()` then it throws the error: `Trainer was configured with "enable_checkpointing=False" but found...

Add Digital typhoon dataset

17

This PR adds the [Digital Typhoon Dataset](http://agora.ex.nii.ac.jp/digital-typhoon/dataset/). The implementation allows the following features: - create an input sequence of single channel images concatenated along channel dimension for nowcasting task (predicting...

Xarray Dataset Support

8

This PR aims to add support for dataloading based on xarray datasets as the discussion began in #1486. The application I have for my project is taking in CMIP6 data...

Fix deterministc group_shuffle_split

1

Sets are unordered and therefore, repeated calls were yielding different train and val sizes for Cyclone dataset.