Learning Multiple Probabilistic Decisions from Latent World Model in Autonomous Driving
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866910617999245312 |
|---|---|
| author | Xiao, Lingyu Liu, Jiang-Jiang Yang, Sen Li, Xiaofan Ye, Xiaoqing Yang, Wankou Wang, Jingdong |
| author_facet | Xiao, Lingyu Liu, Jiang-Jiang Yang, Sen Li, Xiaofan Ye, Xiaoqing Yang, Wankou Wang, Jingdong |
| contents | The autoregressive world model exhibits robust generalization capabilities in vectorized scene understanding but encounters difficulties in deriving actions due to insufficient uncertainty modeling and self-delusion. In this paper, we explore the feasibility of deriving decisions from an autoregressive world model by addressing these challenges through the formulation of multiple probabilistic hypotheses. We propose LatentDriver, a framework models the environment's next states and the ego vehicle's possible actions as a mixture distribution, from which a deterministic control signal is then derived. By incorporating mixture modeling, the stochastic nature of decisionmaking is captured. Additionally, the self-delusion problem is mitigated by providing intermediate actions sampled from a distribution to the world model. Experimental results on the recently released close-loop benchmark Waymax demonstrate that LatentDriver surpasses state-of-the-art reinforcement learning and imitation learning methods, achieving expert-level performance. The code and models will be made available at https://github.com/Sephirex-X/LatentDriver. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2409_15730 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Learning Multiple Probabilistic Decisions from Latent World Model in Autonomous Driving Xiao, Lingyu Liu, Jiang-Jiang Yang, Sen Li, Xiaofan Ye, Xiaoqing Yang, Wankou Wang, Jingdong Robotics Artificial Intelligence The autoregressive world model exhibits robust generalization capabilities in vectorized scene understanding but encounters difficulties in deriving actions due to insufficient uncertainty modeling and self-delusion. In this paper, we explore the feasibility of deriving decisions from an autoregressive world model by addressing these challenges through the formulation of multiple probabilistic hypotheses. We propose LatentDriver, a framework models the environment's next states and the ego vehicle's possible actions as a mixture distribution, from which a deterministic control signal is then derived. By incorporating mixture modeling, the stochastic nature of decisionmaking is captured. Additionally, the self-delusion problem is mitigated by providing intermediate actions sampled from a distribution to the world model. Experimental results on the recently released close-loop benchmark Waymax demonstrate that LatentDriver surpasses state-of-the-art reinforcement learning and imitation learning methods, achieving expert-level performance. The code and models will be made available at https://github.com/Sephirex-X/LatentDriver. |
| title | Learning Multiple Probabilistic Decisions from Latent World Model in Autonomous Driving |
| topic | Robotics Artificial Intelligence |
| url | https://arxiv.org/abs/2409.15730 |