Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2601.18442 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913139102056448 |
|---|---|
| author | Zhao, Hongyi Wang, Shuo He, Qijie Pu, Ziyuan |
| author_facet | Zhao, Hongyi Wang, Shuo He, Qijie Pu, Ziyuan |
| contents | Autonomous Vehicle (AV) requires rigorous testing in safety-critical scenarios for safety validation, yet its validation is hindered by the high cost of field testing and the lack of fidelity in current simulations for rare safety-critical events. Crash reports offer rich and authentic specifications of real-world accident dynamics, making them a promising resource for Large Language Models and Vision-Language models to generate high-fidelity scenarios. However, the existing models frequently deviate from actual accident characteristics due to context suppression. To address these limitations, this paper presents SG-CADVLM, a framework integrateing Context-Aware Decoding with multimodal input processing to generate safety-critical scenarios from crash reports. The framework mitigates the hallucination of VLMs while generating road geometry and vehicle trajectories simultaneously. The experimental results demonstrate that SG-CADVLM generates combined critical and high-risk scenarios at a rate of 88.1% compared to 31.2% for the baseline methods, representing a 182% improvement, while producing executable simulations for autonomous vehicle testing. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_18442 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | SG-CADVLM: A Context-Aware Decoding Powered Vision Language Model for Safety-Critical Scenario Generation Zhao, Hongyi Wang, Shuo He, Qijie Pu, Ziyuan Robotics Autonomous Vehicle (AV) requires rigorous testing in safety-critical scenarios for safety validation, yet its validation is hindered by the high cost of field testing and the lack of fidelity in current simulations for rare safety-critical events. Crash reports offer rich and authentic specifications of real-world accident dynamics, making them a promising resource for Large Language Models and Vision-Language models to generate high-fidelity scenarios. However, the existing models frequently deviate from actual accident characteristics due to context suppression. To address these limitations, this paper presents SG-CADVLM, a framework integrateing Context-Aware Decoding with multimodal input processing to generate safety-critical scenarios from crash reports. The framework mitigates the hallucination of VLMs while generating road geometry and vehicle trajectories simultaneously. The experimental results demonstrate that SG-CADVLM generates combined critical and high-risk scenarios at a rate of 88.1% compared to 31.2% for the baseline methods, representing a 182% improvement, while producing executable simulations for autonomous vehicle testing. |
| title | SG-CADVLM: A Context-Aware Decoding Powered Vision Language Model for Safety-Critical Scenario Generation |
| topic | Robotics |
| url | https://arxiv.org/abs/2601.18442 |