KO | EN

GitLite

Minimalist Open Source Repository Search Engine

gitlite — search

#typescript #ai-agents #cli #claude-code #codex #developer-tools #dsh-plugin #ai-agent #mcp #deepseek-harness #windows #open-source

FollowNet

★ 66 Open GitHub ↗

Source code for paper "FollowNet: A Comprehensive Benchmark for Car-Following Behavior Modeling"

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# FollowNet: A Comprehensive Benchmark for Car-Following Behavior Modeling
Source code for the following paper:

Chen, Xianda, et al. "FollowNet: A Comprehensive Benchmark for Car-Following Behavior Modeling." 
[https://www.nature.com/articles/s41597-023-02718-7](https://www.nature.com/articles/s41597-023-02718-7)

## 📝Description
This notebook demonstrates how to achieve the car following models from traditonal models to data driven models. Motivation: given extracted car following events from five open datasets with the same data formate and train the car follow models. Author: Chen Xianda.

The extracted car following events are avaliable for download.
Provide a tutorial of the data format and how to run the traditional models and the data-driven models.

## 🚕  Data
Extracted car-following events are stored in `data/` folder. The colab tutorial takes the highD data for experiments first. 
     
   The datasets are HighD, SPMD(DAS1, DAS2), Waymo, Lyft, NGSIM. Each has its own training, validation and test part.

## 🛠 Quick Start 
Run the colab notebook directly! Details are in the notebook below.

[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1jB-eM9A1N1q5mPv3TjZPx6drezvURtqD?usp=share_link)

## 📚 Pretrained Models
Pretrained models are stored in `trained_model/` folder.

## 📈 Dataset distribution
Below is the average time gap during car following (s). For more results stored in `results/` folder.

![](https://raw.githubusercontent.com/HKUST-DRIVE-AI-LAB/FollowNet/main/results/time_gap.jpg)

## 📊 Evaluation Metrics
![](https://raw.githubusercontent.com/HKUST-DRIVE-AI-LAB/FollowNet/main/results/tables.png)

Collsion rate

![](https://raw.githubusercontent.com/HKUST-DRIVE-AI-LAB/FollowNet/main/results/collision.png)

MSE of spacing
 
![](https://raw.githubusercontent.com/HKUST-DRIVE-AI-LAB/FollowNet/main/results/mse.png)

## 📭Contact
meixin@ust.hk

xchen595@connect.hkust-gz.edu.cn

## 📎 References

If you use extracted car following data / FollowNet in your own work, please cite:

```latex
Chen, X., Zhu, M., Chen, K. et al. FollowNet: A Comprehensive Benchmark for Car-Following Behavior Modeling. Sci Data 10, 828 (2023). https://doi.org/10.1038/s41597-023-02718-7

```

FollowNet

Explore Similar Repositories

Was this content helpful?

Recent Feedback

Download README