MetaUrban: A Simulation Platform for Embodied AI in Urban Spaces

[A tiny version for review]

🛠 Quick Start

Install MetaUrban (Tiny Version) via:

git clone -b tiny https://github.com/metadriverse/metaurban
cd metaurban
pip install -e .

download assets from

https://drive.google.com/file/d/194pgea_J7mjjlmFD4pzj3KQAWxsJtmeL/view?usp=sharing

unzip the file and organize the folder as

-metaurban
  -metaurban
      -assets
      -assets_pedestrian
      -base_class
      -...

install ORCA algorithm for trajectory generation

conda install pybind11 -c conda-forge
pip install scikit-image
cd metaurban/orca_algo
rm -rf build
bash compile.sh 

install torch and stable-baselines3 for RL training

pip install torch
pip install stable_baselines3

we put a small asset subset of objects and agents under the folder metaurban/assets/models/test/ and metaurban/assets_pedestrian/. You should change paths in path_config.yaml.

metaurbanasset: /PATH/TO/ASSETS

parentfolder: /PATH/TO/AdjustedParameters

Note that the program is tested on Linux, Windows and WSL2. Some issues in MacOS wait to be solved.

MetaUrban Simulator Roam

We provide examples to demonstrate features and basic usages of metaurban after the local installation.

Point Navigation Environment

In point navigation environment, there will be only static objects besides the ego agent in the scenario.

Run the following command to launch a simple scenario with manual control. Press W,S,A,D to control the delivery robot.

python -m metaurban.examples.drive_in_static_env

Press key R for loading a new scenario. If there is no response when you press W,S,A,D, press T to enable manual control.

Social Navigation Environment

In social navigation environment, there will be vehicles, pedestrians and some other agents in the scenario.

Run the following command to launch a simple scenario with manual control. Press W,S,A,D to control the delivery robot.

python -m metaurban.examples.drive_in_dynamic_env

MetaUrban-12K Dataset Generation

We provide a subset of seeds with selected ORCA reference trajectory for ego agent. You can run the command as below to generate a scenario from these seeds

python -m metaurban.scripts.generate_static_scenario

for PointNav environment

python -m metaurban.scripts.generate_dynamic_scenario

for SocialNav environment

The file valid_seed.pkl in the root folder is a subset of MetaUrban-12K Dataset. It's notable that currently provided version do not match the version we use since complete asset library is not provided yet, there may still be some unreasonable reference trajectories in the scenario.