探索模拟城市中车辆导航的多模态隐式行为学习
Exploring multimodal implicit behavior learning for vehicle navigation in simulated cities
AI总结:
针对标准行为克隆无法学习多模态驾驶决策的问题,本文提出数据增强的隐式行为克隆(DA-IBC),通过扰动专家动作构造反例并改进无导数推理初始化,在CARLA城市驾驶任务中有效学习多模态动作分布并优于标准IBC。
摘要:
标准行为克隆(BC)无法学习多模态驾驶决策,因为在同一场景下存在多个有效动作。我们探索了基于能量模型(EBMs)的隐式行为克隆(IBC),以更好地捕捉这种多模态性。我们提出了数据增强的隐式行为克隆(DA-IBC),该方法通过对专家动作进行扰动以形成IBC训练的反例,并为无导数推理使用更好的初始化,从而改进学习效果。在CARLA模拟器中以鸟瞰图(Bird's-Eye View)作为输入的实验表明,在测试环境中为评估多模态行为学习而设计的城市驾驶任务中,DA-IBC优于标准IBC。学习到的能量景观能够表示多模态动作分布,而BC无法实现这一点。
英文摘要:
Standard Behavior Cloning (BC) fails to learn multimodal driving decisions, where multiple valid actions exist for the same scenario. We explore Implicit Behavioral Cloning (IBC) with Energy-Based Models (EBMs) to better capture this multimodality. We propose Data-Augmented IBC (DA-IBC), which improves learning by perturbing expert actions to form the counterexamples of IBC training and using better initialization for derivative-free inference. Experiments in the CARLA simulator with Bird's-Eye View inputs demonstrate that DA-IBC outperforms standard IBC in urban driving tasks designed to evaluate multimodal behavior learning in a test environment. The learned energy landscapes are able to represent multimodal action distributions, which BC fails to achieve.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。