Fine-Tuned UNet with resnext50_32x4d backbone for Building Damage Assessment on LA fire of January 2025
This repository contains the fine-tuned weights for a UNet with resnext50_32x4d backbone (default architecture of Microsoft's damage-building-assessment repository), to perform building damage assessment. The model takes an RGB image as input and outputs a segmented image where each pixel represents one of three classes: background, healthy building, or damaged building (example below).
How to Use the Fine-Tuned Weights
Follow the steps below to use the fine-tuned weights and perform building damage assessment on your custom images:
1. Download the Fine-Tuned Weights
Download the fine-tuned weights from this repo
2. Clone the Building Damage Assessment Repository
Clone the damage-building-assessment repository.
3. Create a new experiment
Create a new config file for your experiment in the config folder and run the script project_setup.py
4. Add the weights
Add the fine-tuned weights in the experiments/your-experiment/checkpoints folder
5. Run inference.py
run the script inference.py on your custom images.
More instructuons are provided here