Skip to content

Latest commit

 

History

History
49 lines (38 loc) · 1.06 KB

File metadata and controls

49 lines (38 loc) · 1.06 KB

Arm-UNICEF-Disaster-Vulnerability

Identifying thatch-roofed houses for disaster planning in rural Malawi using aerial images

Description

  • Over 80% of Malawi's population resides in rural areas.
  • Natural disasters and global challenges like Covid-19 affect these communities.
  • Current damage assessment methods overlook crucial data, such as identifying houses with grass-thatched roofs.
  • This competition leverages machine learning on aerial imagery to accurately count these structures.

Set Up

Create a virtual environment

python -m venv venv
source /venv/bin/activate

Install requirements

pip install -r requirements.txt

Resize Images

python resize_images.py

Prepare data for training in Yolo format

python prepare_data.py

Augment data

python data_aug.py

Train models

python trainYolo.py

Dataset

Download the dataset

You can download the dataset from here

Example image

Labeled houses