Update README.md
Browse files
README.md
CHANGED
|
@@ -7,4 +7,55 @@ language:
|
|
| 7 |
tags:
|
| 8 |
- Autonomous-Driving
|
| 9 |
- Simulation
|
| 10 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 7 |
tags:
|
| 8 |
- Autonomous-Driving
|
| 9 |
- Simulation
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
<div id="top" align="center">
|
| 13 |
+
|
| 14 |
+
<p align="center">
|
| 15 |
+
<img src="https://ik.imagekit.io/StarBurger/SimScale/title_1080p.gif">
|
| 16 |
+
</p>
|
| 17 |
+
|
| 18 |
+
# **Learning to Drive via Real-World Simulation at Scale**
|
| 19 |
+
|
| 20 |
+
[](https://arxiv.org/abs/2511.23369)
|
| 21 |
+
[](https://opendrivelab.com/SimScale/)
|
| 22 |
+
[](https://huggingface.co/datasets/OpenDriveLab/SimScale)
|
| 23 |
+
[](https://modelscope.cn/datasets/OpenDriveLab/SimScale)
|
| 24 |
+
[](https://github.com/OpenDriveLab/SimScale/blob/main/LICENSE)
|
| 25 |
+
|
| 26 |
+
</div>
|
| 27 |
+
|
| 28 |
+
<div id="top" align="center">
|
| 29 |
+
<p align="center">
|
| 30 |
+
<img src="assets/teaser.png" >
|
| 31 |
+
</p>
|
| 32 |
+
</div>
|
| 33 |
+
|
| 34 |
+
|
| 35 |
+
> [Haochen Tian](https://github.com/hctian713),
|
| 36 |
+
> [Tianyu Li](https://github.com/sephyli),
|
| 37 |
+
> [Haochen Liu](https://georgeliu233.github.io/),
|
| 38 |
+
> [Jiazhi Yang](https://github.com/YTEP-ZHI),
|
| 39 |
+
> [Yihang Qiu](https://github.com/gihharwtw),
|
| 40 |
+
> [Guang Li](https://scholar.google.com/citations?user=McEfO8UAAAAJ&hl=en),
|
| 41 |
+
> [Junli Wang](https://openreview.net/profile?id=%7EJunli_Wang4),
|
| 42 |
+
> [Yinfeng Gao](https://scholar.google.com/citations?user=VTn0hqIAAAAJ&hl=en),
|
| 43 |
+
> [Zhang Zhang](https://scholar.google.com/citations?user=rnRNwEMAAAAJ&hl=en),
|
| 44 |
+
> [Liang Wang](https://scholar.google.com/citations?user=8kzzUboAAAAJ&hl=en),
|
| 45 |
+
> [Hangjun Ye](https://scholar.google.com/citations?user=68tXhe8AAAAJ&hl=en),
|
| 46 |
+
> [Tieniu Tan](https://scholar.google.com/citations?user=W-FGd_UAAAAJ&hl=en),
|
| 47 |
+
> [Long Chen](https://long.ooo/),
|
| 48 |
+
> [Hongyang Li](https://lihongyang.info/)
|
| 49 |
+
>
|
| 50 |
+
>
|
| 51 |
+
> - π§ Primary Contact: Haochen Tian (tianhaochen2023@ia.ac.cn)
|
| 52 |
+
> - π Materials: π [π](https://x.com/OpenDriveLab/status/1999507869633527845) | π° [Media](https://mp.weixin.qq.com/s/OGV3Xlb0bHSSSloG11qFJA) | ποΈ [Slides](https://docs.google.com/presentation/d/17qbsKZU9jdw7MfiPk7hZelaLb3leR2M76gPcMkuf1MI/edit?usp=sharing) | π¬ [Talk (in Chinese)](https://www.bilibili.com/video/BV1tqrEBNECQ)
|
| 53 |
+
> - ποΈ Joint effort by CASIA, OpenDriveLab at HKU, and Xiaomi EV.
|
| 54 |
+
|
| 55 |
+
---
|
| 56 |
+
|
| 57 |
+
## π₯ Highlights
|
| 58 |
+
|
| 59 |
+
- ποΈ A scalable simulation pipepline that synthesizes diverse and high-fidelity reactive driving scenarios with pseudo-expert demonstrations.
|
| 60 |
+
- π An effective sim-real co-training strategy that improves robustness and generalization synergistically across various end-to-end planners.
|
| 61 |
+
- π¬ A comprehensive recipe that reveals crucial insights into the underlying scaling properties of sim-real learning systems for end-to-end autonomy.
|