README.md

metadata

license: apache-2.0
language:
  - en
tags:
  - Pytorch
  - mmsegmentation
  - segmentation
  - Crop Classification
  - Multi Temporal
  - Geospatial
  - Foundation model
datasets:
  - ibm-nasa-geospatial/hls_burn_scars
metrics:
  - accuracy
  - IoU

Model and Inputs

The pretrained Prithvi-100m parameter model is finetuned to classify crop and other land cover types based off HLS data and CDL labels from the HLS Burn Scar Scenes dataset. This dataset includes input chips of 224x224x18, where 224 is the height and width and 18 is combined with 6 bands of 3 time-steps. The bands are:

1. Blue
2. Green
3. Red
4. Narrow NIR
5. SWIR 1
6. SWIR 2

While the Prithvi-100m was pretrained with 3 timesteps, this task utilize the capibility of multi-temporal data input adapted from the pretrained foundation model and provide more generalized and

Code

Code for Finetuning is available through github

Configuration used for finetuning is available through config.

Results

The experiment by running the mmseg stack for 80 epochs using the above config led to an IoU of 0.72 on the burn scar class and 0.96 overall accuracy. It is noteworthy that this leads to a resonably good model, but further developement will most likely improve performance.

Inference and demo

There is an inference script that allows to run the hls-cdl crop classification model for inference on HLS images. These input have to be geotiff format, including 18 bands for 3 time-step, and each time-step includes the channels described above (Blue, Green, Red, Narrow NIR, SWIR, SWIR 2) in order. There is also a demo that leverages the same code here.