Generating Realistic Safety-Critical Scenarios for Vehicle-Pedestrian Interactions
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866910228509884416 |
|---|---|
| author | Pu, Qingwen Xie, Kun Zhu, Yuan Zhai, Guocong |
| author_facet | Pu, Qingwen Xie, Kun Zhu, Yuan Zhai, Guocong |
| contents | Automated driving system deployment requires rigorous validation across safety-critical vehicle-pedestrian interactions, yet real-world datasets rarely capture high-risk scenarios while simulation platforms lack realistic behavior. In response, this study proposes a three-stage framework that combines real-world grounding with adaptive simulation to generate behaviorally realistic safety-critical scenarios at scale. Stage 1 pre-trains multi-agent state-space Transformer-enhanced DDPG (MA-SST-DDPG) agents on real-world safety-critical data to learn human-like interactive evasive behaviors through data-driven learning. Stage 2 deploys pre-trained multi-agents in CARLA for online reinforcement learning to generalize across diverse scenarios, integrating real-world knowledge with simulation experience to produce a refined MA-SST-DDPG model. Stage 3 uses CARLA with the refined model to generate over 198,000 high-resolution interaction episodes from eight intersection scenarios, culminating in the Vehicle-Pedestrian Safety-Critical Interaction (VPSCI) dataset. The Refined MA-SST-DDPG model outperformed baseline methods in reproducing realistic evasive behaviors, achieving the lowest trajectory errors (ADE = 0.072 m, FDE = 0.142 m). Statistical comparison confirmed distributional equivalence between the generated and real-world data in both conflict severity and behavioral response. A Turing test confirmed that the three-stage framework generated evasive behaviors were indistinguishable from real-world interactions. These results demonstrate the framework's effectiveness in producing high-fidelity safety-critical data, offering valuable sources for the development of ADS and simulation-based safety evaluations. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_17229 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Generating Realistic Safety-Critical Scenarios for Vehicle-Pedestrian Interactions Pu, Qingwen Xie, Kun Zhu, Yuan Zhai, Guocong Robotics Systems and Control 68T05, 68T42, 68T07, 91A15, 90B20 I.2.11; I.2.6; I.6.3; I.2.9 Automated driving system deployment requires rigorous validation across safety-critical vehicle-pedestrian interactions, yet real-world datasets rarely capture high-risk scenarios while simulation platforms lack realistic behavior. In response, this study proposes a three-stage framework that combines real-world grounding with adaptive simulation to generate behaviorally realistic safety-critical scenarios at scale. Stage 1 pre-trains multi-agent state-space Transformer-enhanced DDPG (MA-SST-DDPG) agents on real-world safety-critical data to learn human-like interactive evasive behaviors through data-driven learning. Stage 2 deploys pre-trained multi-agents in CARLA for online reinforcement learning to generalize across diverse scenarios, integrating real-world knowledge with simulation experience to produce a refined MA-SST-DDPG model. Stage 3 uses CARLA with the refined model to generate over 198,000 high-resolution interaction episodes from eight intersection scenarios, culminating in the Vehicle-Pedestrian Safety-Critical Interaction (VPSCI) dataset. The Refined MA-SST-DDPG model outperformed baseline methods in reproducing realistic evasive behaviors, achieving the lowest trajectory errors (ADE = 0.072 m, FDE = 0.142 m). Statistical comparison confirmed distributional equivalence between the generated and real-world data in both conflict severity and behavioral response. A Turing test confirmed that the three-stage framework generated evasive behaviors were indistinguishable from real-world interactions. These results demonstrate the framework's effectiveness in producing high-fidelity safety-critical data, offering valuable sources for the development of ADS and simulation-based safety evaluations. |
| title | Generating Realistic Safety-Critical Scenarios for Vehicle-Pedestrian Interactions |
| topic | Robotics Systems and Control 68T05, 68T42, 68T07, 91A15, 90B20 I.2.11; I.2.6; I.6.3; I.2.9 |
| url | https://arxiv.org/abs/2605.17229 |