This repo contains code needed to create the training data and train an image segmentation model to recognise certain parts of floorplans. We apply these trained models to a dataset of floor plans in order to estimate how many rooms, different room types, windows and doors they have.
Please refer to the README in the pipeline folder for more information about the methodology, usage and results.
direnv
and conda
make install
to configure the development environment:
pre-commit
export METAFLOW_PROFILE=floorplan
and direnv reload
METAFLOW_USER
name has no fullstops in.Technical and working style guidelines
Project based on Nesta's data science project template (Read the docs here).