Skip to content

Adelaide-Autonomous-Racing-Kit/ac-segmentation

Repository files navigation

Semantic Segmentation using ACDG Data

Setup

w&b

Export your w&b API key prior to running a training job using: export WANDB_API_KEY=<your api key>

Docker

Modify the docker/compose.yaml file to mount your data to the container and change any training settings you wish.`

Data

You can use semantic segment generated by ACDG from your own ACI recordings. Alternatively, the set of training data used to train models referenced in our paper can be downloaded here

Run

A training job can be run in docker using the command bash train.sh.

Citation

If you use this code, data or models in your research please cite us in your work:

@misc{bockman2024aarkopentoolkitautonomous,
      title={AARK: An Open Toolkit for Autonomous Racing Research}, 
      author={James Bockman and Matthew Howe and Adrian Orenstein and Feras Dayoub},
      year={2024},
      eprint={2410.00358},
      archivePrefix={arXiv},
      primaryClass={cs.RO},
      url={https://arxiv.org/abs/2410.00358}, 
}

About

Training code for segmentation models used in ac-mpc

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Contributors 3

  •  
  •  
  •