Simulanes

CARLA synthetic dataset generator for the paper Sim-to-Real Domain Adaptation for Lane Detection and Classification in Autonomous Driving. If you use our dataset generator or our simulation dataset, please cite our paper:

@misc{hu2022simtoreal,
      title={Sim-to-Real Domain Adaptation for Lane Detection and Classification in Autonomous Driving}, 
      author={Chuqing Hu and Sinclair Hudson and Martin Ethier and Mohammad Al-Sharman and Derek Rayside and William Melek},
      year={2022},
      eprint={2202.07133},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}

Usage

Carla server

Install docker and nvidia-container-runtime. Then pull the carla docker image:

docker pull carlasim/carla:0.9.11

Build the server image (downloads additional assets):

bash docker/server/build_server.sh

Run the server container:

bash docker/server/run_server.sh

Carla client

Build the client image:

bash docker/client/build_client.sh

Run the client container:

bash docker/client/run_client.sh

Run the Python script from within the client container:

python3.7 automatic_control.py

usage: automatic_control.py [-h] [-v] [--host H] [-p P] [--res WIDTHxHEIGHT]
                            [--filter PATTERN] [--gamma GAMMA]
                            [-b {cautious,normal,aggressive}]
                            [-a {Behavior,Roaming,Basic}] [-s SEED]
                            [-i NUM_IMAGES] [-r] [-n N] [-w W] [--safe]
                            [--filterv PATTERN] [--filterw PATTERN]
                            [--tm_port P] [--sync] [--hybrid]
                            [--car_lights_on]

CARLA Automatic Control Client

optional arguments:
  -h, --help            show this help message and exit
  -v, --verbose         Print debug information
  --host H              IP of the host server (default: 127.0.0.1)
  -p P, --port P        TCP port to listen to (default: 2000)
  --res WIDTHxHEIGHT    Window resolution (default: 1280x720)
  --filter PATTERN      Actor filter (default: "vehicle.tesla.model3")
  --gamma GAMMA         Gamma correction of the camera (default: 2.2)
  -b {cautious,normal,aggressive}, --behavior {cautious,normal,aggressive}
                        Choose one of the possible agent behaviors (default:
                        normal)
  -a {Behavior,Roaming,Basic}, --agent {Behavior,Roaming,Basic}
                        select which agent to run
  -s SEED, --seed SEED  Set seed for repeating executions (default: None)
  -i NUM_IMAGES, --num_images NUM_IMAGES
                        Set total number of image to generate (default: 1000)
  -r, --resume          Continue from previous session resume_progress.json
                        file
  -n N, --num_npc_vehicles N
                        Max number of NPC vehicles (default: 20)
  -w W, --num_npc_walkers W
                        Max number of NPC walkers (default: 10)
  --safe                Avoid spawning vehicles prone to accidents
  --filterv PATTERN     Vehicles filter (default: "vehicle.*")
  --filterw PATTERN     Pedestrians filter (default: "walker.pedestrian.*")
  --tm_port P           Port to communicate with TM (default: 8000)
  --sync                Synchronous mode execution
  --hybrid              Enable hybrid physics mode
  --car_lights_on       Enable car lights at spawn

For manual control, run this Python script instead:

python3.7 manual_control.py [-t TOWN]

See src/README.md for details on our code such as

• Data format
• Adding new maps
• Dynamic weather configuration

Reference

Our code is built upon PythonAPI/examples from carla-simulator/carla.

If it is helpful for your work, please cite:

CARLA: An Open Urban Driving Simulator
Alexey Dosovitskiy, German Ros, Felipe Codevilla, Antonio Lopez, Vladlen Koltun; PMLR 78:1-16 [PDF] [talk]

@inproceedings{Dosovitskiy17,
  title = {{CARLA}: {An} Open Urban Driving Simulator},
  author = {Alexey Dosovitskiy and German Ros and Felipe Codevilla and Antonio Lopez and Vladlen Koltun},
  booktitle = {Proceedings of the 1st Annual Conference on Robot Learning},
  pages = {1--16},
  year = {2017}
}