GenAD: Generative End-to-End Autonomous Driving

January 11, 2025 · View on GitHub

Paper

GenAD: Generative End-to-End Autonomous Driving

Wenzhao Zheng*, Ruiqi Song*, Xianda Guo* †\dagger, Chenming Zhang, Long Chen†\dagger

* Equal contributions †\dagger Corresponding authors

GenAD casts autonomous driving as a generative modeling problem.

News

  • [2024/11/10] Closed-loop code for GenAD has been released.
  • [2024/5/2] Training and evaluation code release.
  • [2024/2/18] Paper released on arXiv.

Demo

demo

Overview

comparison

Comparisons of the proposed generative end-to-end autonomous driving framework with the conventional pipeline. Most existing methods follow a serial design of perception, prediction, and planning. They usually ignore the high-level interactions between the ego car and other agents and the structural prior of realistic trajectories. We model autonomous driving as a future generation problem and conduct motion prediction and ego planning simultaneously in a structural latent trajectory space.

Closed-Loop Autonomous Driving

vis

Bench2Drive comprises the Bench2Drive repository for closed-loop evaluation and the model repository Bench2DriveZoo. The code in this repository integrates GenAD within the Bench2DriveZoo repository, with the majority of the code being identical to that in Bench2DriveZoo. This repository does not contain the code from the Bench2Drive repository, and no modifications were made to the closed-loop evaluation code. Only the execution scripts were adjusted, as detailed in the following description.

Results

Open-Loop Setting on nuScenes

results

Closed-Loop Setting on Carla

MethodDriving ScoreSuccess Rates (%)
VAD (Paper)39.420.1
VAD (Github Update)42.350.13
VAD (Reproduction)38.160.15
GenAD44.810.159

Open-Loop Code

Dataset

Download nuScenes V1.0 full dataset data and CAN bus expansion data HERE. Prepare nuscenes data as follows.

Download CAN bus expansion

# download 'can_bus.zip'
unzip can_bus.zip 
# move can_bus to data dir

Prepare nuScenes data

We genetate custom annotation files which are different from mmdet3d's

Generate the train file and val file:

python tools/data_converter/genad_nuscenes_converter.py nuscenes --root-path ./data/nuscenes --out-dir ./data/nuscenes --extra-tag genad_nuscenes --version v1.0 --canbus ./data

Using the above code will generate genad_nuscenes_infos_temporal_{train,val}.pkl.

Folder structure

GenAD
├── projects/
├── tools/
├── configs/
├── ckpts/
│   ├── resnet50-19c8e357.pth
├── data/
│   ├── can_bus/
│   ├── nuscenes/
│   │   ├── maps/
│   │   ├── samples/
│   │   ├── sweeps/
│   │   ├── v1.0-test/
|   |   ├── v1.0-trainval/
|   |   ├── genad_nuscenes_infos_train.pkl
|   |   ├── genad_nuscenes_infos_val.pkl

Installation

Detailed package versions can be found in requirements.txt.

Open-Loop Evaluation

datasets

https://drive.google.com/drive/folders/1gy7Ux-bk0sge77CsGgeEzPF9ImVn-WgJ?usp=drive_link

Checkpoints

https://drive.google.com/drive/folders/1nlAWJlvSHwqnTjEwlfiE99YJVRFKmqF9?usp=drive_link

Train GenAD with 8 GPUs

cd /path/to/GenAD
conda activate genad
python -m torch.distributed.run --nproc_per_node=8 --master_port=2333 tools/train.py projects/configs/GenAD/GenAD_config.py --launcher pytorch --deterministic --work-dir path/to/save/outputs

Eval GenAD with 1 GPU

cd /path/to/GenAD
conda activate genad
CUDA_VISIBLE_DEVICES=0 python tools/test.py projects/configs/VAD/GenAD_config.py /path/to/ckpt.pth --launcher none --eval bbox --tmpdir outputs

Closed-Loop Code

Installation

Clone this repository and configure it according to the Getting Started section in the Bench2DriveZoo repository documentation. Refer to the configuration documentation in the Bench2Drive repository to link this repository to the closed-loop evaluation repository.

Detailed package versions can be found in requirements.txt.

Training

sh ./adzoo/genad/dist_train.sh ./adzoo/genad/configs/VAD/GenAD_config_b2d.py 1

Note: Detailed training and evaluation methods can be found in the documentation of Bench2DriveZoo.

Open-Loop Evaluation

sh ./adzoo/genad/dist_test.sh ./adzoo/genad/configs/VAD/GenAD_config_b2d.py ./work_dirs/GenAD_config_b2d/epoch_.pth 1

Closed-Loop Evaluation

Eval GenAD with 8 GPUs

leaderboard/scripts/run_evaluation_multi.sh

Eval GenAD with 1 GPU

leaderboard/scripts/run_evaluation_debug.sh

Note: Detailed training and evaluation methods can be found in the documentation of Bench2DriveZoo.

Our code is based on VAD and UniAD.

Closed-loop code is based on Bench2DriveZoo.

Citation

If you find this project helpful, please consider citing the following paper:

@article{zheng2024genad,
    title={GenAD: Generative End-to-End Autonomous Driving},
    author={Zheng, Wenzhao and Song, Ruiqi and Guo, Xianda and Zhang, Chenming and Chen, Long},
    journal={arXiv preprint arXiv: 2402.11502},
    year={2024}
}