environment-generator-3d

A dynamic indoor 3D environment generator for robot learning and QD.

Project Setup and Run Instructions

Follow the following steps to setup the repository locally:

• Clone the repo: git clone git@github.com:l0g1c-80m8/environment-generator-3d.git.
• Create a new python environment and install required dependencies using pip install requirements.txt.
• Create a new directory called dataset in the project root directory.
• Download the 3D-Front and 3D-Future datasets and extract them in the dataset directory.
• Create a new directory called out in the project root directory.
• Execute preprocess script to generate all artefacts before running simulations using ./preprocess.sh ../dataset/3D-FRONT ../dataset/3D-FUTURE-model ../out categories.json furniture.json furniture_bbox.json from the preprocess dir.
• Launch the simulation by navigation to src directory and running python main.py.

The project structure should look like this: