File size: 2,331 Bytes
7c4494a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
---
task_categories:
- robotics
---

# Fail2Drive: Benchmarking Closed-Loop Driving Generalization

[**Project Page**](https://simonger.github.io/fail2drive/) | [**Paper**](https://huggingface.co/papers/2604.08535) | [**GitHub**](https://github.com/autonomousvision/fail2drive)

Fail2Drive is the first CARLA v2 benchmark designed to test closed-loop generalization on truly unseen long-tail scenarios. By pairing each shifted route with an in-distribution reference scenario, it exposes substantial hidden failure modes in current state-of-the-art driving models.

## Highlights
- **17 unseen scenarios** for evaluation of true generalization.
- **30 novel assets** including animals, visual noise, and adversarial obstacles.
- **Paired route design** enables quantification of the generalization gap.
- **100 route pairs** in diverse environments and configurations.
- **Toolbox** for creating custom obstacles and routes.

## Installation

To set up the environment and the Fail2Drive CARLA simulator, follow these steps:

```bash
# 1. Clone this repository
git clone https://github.com/autonomousvision/fail2drive.git
cd fail2drive

# 2. Set up the Fail2Drive CARLA simulator
mkdir f2d_carla
curl -L \
  https://huggingface.co/datasets/SimonGer/fail2drive/resolve/main/fail2drive_simulator.tar.gz \
  | tar -xz -C f2d_carla

# 3. Create the conda environment
conda env create -f environment.yml
conda activate fail2drive

# 4. Set environment variables
source env_vars.sh
```

## Sample Usage

To run a keyboard-controlled human agent on a benchmark route, start CARLA in a separate terminal and run:

```bash
python leaderboard/leaderboard/leaderboard_evaluator.py \
  --agent ${WORK_DIR}/leaderboard/leaderboard/autoagents/human_agent_keyboard.py \
  --routes ${WORK_DIR}/fail2drive_split/Generalization_PedestriansOnRoad_1085.xml
```

To run the PDM-Lite expert policy:

```bash
python leaderboard/leaderboard/leaderboard_evaluator_local.py \
  --agent ${WORK_DIR}/team_code/visu_agent.py \
  --track MAP \
  --routes ${WORK_DIR}/fail2drive_split/Generalization_PedestriansOnRoad_1085.xml
```

## Citation

```bibtex
@article{gerstner2024fail2drive,
  title={Fail2Drive: Benchmarking Closed-Loop Driving Generalization},
  author={Gerstner, Simon and others},
  journal={arXiv preprint arXiv:2604.08535},
  year={2024}
}
```