README.md

metadata

license: apache-2.0
datasets:
  - jonathan-roberts1/Satellite-Images-of-Hurricane-Damage
tags:
  - climate

Model Card for fluffypotato/flooding-classifier

This model is trained on a subset of jonathan-roberts1/Satellite-Images-of-Hurricane-Damage? and performs binary classification on satellite images into either damaged or not damaged.

The model was trained using PyTorch on Intel Developer Cloud using ipex optimization. Here is the Neural Network architecture:

model.fc = nn.Sequential(
           nn.Linear(2048, 128),
           nn.ReLU(inplace=True),
            nn.Linear(128, 64),
           nn.ReLU(inplace=True),
           nn.Linear(64, 2))

Uses

Some use cases include emergency response and disaster detection.

Bias, Risks, and Limitations

This model is limited in use as it only performs binary classification. Further extensions include adding more classes to the model or more diverse data to the model.