Adv-BMT: Bidirectional Motion Transformer for Safety-Critical Traffic Scenario Generation
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866912686267170816 |
|---|---|
| author | Liu, Yuxin Peng, Zhenghao Cui, Xuanhao Zhou, Bolei |
| author_facet | Liu, Yuxin Peng, Zhenghao Cui, Xuanhao Zhou, Bolei |
| contents | Scenario-based testing is essential for validating the performance of autonomous driving (AD) systems. However, such testing is limited by the scarcity of long-tailed, safety-critical scenarios in existing datasets collected in the real world. To tackle the data issue, we propose the Adv-BMT framework, which augments real-world scenarios with diverse and realistic adversarial traffic interactions. The core component of Adv-BMT is a bidirectional motion transformer (BMT) model to perform inverse traffic motion predictions, which takes agent information in the last time step of the scenario as input, and reconstructs the traffic in the inverse of chronological order until the initial time step. The Adv-BMT framework is a two-staged pipeline: it first conducts adversarial initializations and then inverse motion predictions. Different from previous work, we do not need any collision data for pretraining, and are able to generate realistic and diverse collision interactions. Our experimental results validate the quality of generated collision scenarios by Adv-BMT: training in our augmented dataset would reduce episode collision rates by 20%. Demo and code are available at: https://metadriverse.github.io/adv-bmt/. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_09485 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Adv-BMT: Bidirectional Motion Transformer for Safety-Critical Traffic Scenario Generation Liu, Yuxin Peng, Zhenghao Cui, Xuanhao Zhou, Bolei Robotics Artificial Intelligence Graphics Scenario-based testing is essential for validating the performance of autonomous driving (AD) systems. However, such testing is limited by the scarcity of long-tailed, safety-critical scenarios in existing datasets collected in the real world. To tackle the data issue, we propose the Adv-BMT framework, which augments real-world scenarios with diverse and realistic adversarial traffic interactions. The core component of Adv-BMT is a bidirectional motion transformer (BMT) model to perform inverse traffic motion predictions, which takes agent information in the last time step of the scenario as input, and reconstructs the traffic in the inverse of chronological order until the initial time step. The Adv-BMT framework is a two-staged pipeline: it first conducts adversarial initializations and then inverse motion predictions. Different from previous work, we do not need any collision data for pretraining, and are able to generate realistic and diverse collision interactions. Our experimental results validate the quality of generated collision scenarios by Adv-BMT: training in our augmented dataset would reduce episode collision rates by 20%. Demo and code are available at: https://metadriverse.github.io/adv-bmt/. |
| title | Adv-BMT: Bidirectional Motion Transformer for Safety-Critical Traffic Scenario Generation |
| topic | Robotics Artificial Intelligence Graphics |
| url | https://arxiv.org/abs/2506.09485 |