This repository contains all code and resources for experiments conducted in our paper, U-Prithvi: Integrating a GeoAI Foundation Model with UNet for Flood Inundation Mapping. Our research introduces U-Prithvi, a novel framework that combines a GeoAI foundation model with a UNet architecture to enhance flood inundation mapping.

• Python 3.12

1. Clone the repository

   git clone https://github.com/your-repo/segmentation-floods.git
   cd segmentation-floods

2. Install all required packages

   pip install -r requirements.txt

3. Download the Prithvi model

   # Ensure git-lfs is installed (https://git-lfs.com)
   git lfs install
   git clone https://huggingface.co/ibm-nasa-geospatial/Prithvi-100M
   # Rename to a valid Python module name
   mv Prithvi-100M prithvi
   touch prithvi/__init__.py

4. Download the Sen1Floods11 dataset

   Follow the download instructions here. The folder structure should be as follows:

   sen1floods11/
   ├── LabelHand/
   ├── S1Hand/
   ├── S2Hand/
   └── splits/
   

To train the model, execute the following command:

python train.py *params*

This will start the training process using the specified dataset and model directory.

• --data_dir: Path to the dataset directory.
• --model_dir: Path to the Prithvi model directory.
• --batch_size: Batch size for training (default: 32).
• --epochs: Number of epochs for training (default: 50).
• --learning_rate: Learning rate for the optimizer (default: 0.001).
• --checkpoint_dir: Directory to save model checkpoints.
• --log_dir: Directory to save training logs.

• --combine_func: Combination function applied only for U-Prithvi mode. Options: [concat, product, sum]. (default: concat)
• --random_dropout_prob: The probability that one of the embeddings will be dropped. (default: 2/3)