torchgeo
torchgeo copied to clipboard
microsoft
TorchGeo: datasets, samplers, transforms, and pre-trained models for geospatial data
Bumps [rasterio](https://github.com/rasterio/rasterio) from 1.3.0 to 1.3.0.post1. Release notes Sourced from rasterio's releases. 1.3.0.post1 This release fixes two packaging issues. The project code itself has no changes other than an updated...
![dependabot[bot] avatar](https://avatars.githubusercontent.com/in/29110?v=4 "dependabot[bot] avatar"){kind=avatar}
### Description @calebrob6 flagged an issue in merging #510 that the `utils.download_radiant_mlhub_collection(...)` method is resulting in a `FileNotFoundError`. This is because the method in `torchgeo.utils` is expecting the `radiant_mlhub.Dataset.download(...)` method...
Original authors got RMSE: 0.657, MAE: 0.248 with ResCeption, I got RMSE: 0.597, MAE: 0.290 with ResNet-50 Optimal hyperparams are pre-trained ResNet-50 with a learning rate of 0.0001. Still need...
I believe this was a holdover from 0.1.0 back when `IntersectionDataset` didn't have its own index. Current best practice is to use `IntersectionDataset` instead of one of the individual datasets...
### Issue Hello, In torchgeo.model the resnet50 model pretrained on Sentinel-2 data uses 10 channels , but I cannot find anywhere which of the 13 channels are used and in...
### Summary Our current `train.py` script only logs to TensorBoard. We should add a CSV logger. ### Rationale In order to determine the optimal hyperparams in a hyperparam sweep, we...
This PR adds the [Cloud Cover Detection Challenge ](https://mlhub.earth/data/ref_cloud_cover_detection_challenge_v1) dataset which was generated as part of a [crowdsourcing competition](https://www.drivendata.org/competitions/83/cloud-cover/) as hosted on the Radiant MLHub. Dataset features: * Sentinel-2 L2A...
`RasterDataset` by default loads all bands. This PR allows us to be more selective.
### Summary Is there a plan to add the [GeoLifeCLEF dataset ](https://www.kaggle.com/competitions/geolifeclef-2022-lifeclef-2022-fgvc9/data)to the current NonGeoDataset? ### Rationale it is used in well established machine learning problem to address species distribution...
