论文标题
TAE：半监督的可控制行为感知的轨迹发生器和预测变量
TAE: A Semi-supervised Controllable Behavior-aware Trajectory Generator and Predictor
论文作者
论文摘要
轨迹产生和预测是两个交织的任务，在计划者评估和智能车辆决策中起着重要作用。大多数现有方法都集中在两者中之一，并经过优化，可以直接输出最终生成/预测的轨迹，该轨迹仅包含有限的信息，以增加关键场景增强和安全计划。在这项工作中，我们提出了一种新型的行为感知的轨迹自动编码器（TAE），该自动编码器（TAE）使用半监督的对抗自动编码器和运输中的域知识来明确对驱动因素的行为进行建模，例如侵略性和潜在空间中的侵略性和意图。我们的模型解决了统一体系结构和益处的轨迹产生和预测：该模型可以生成多样化，可控和现实的轨迹，以增强计划者在安全至关重要和长尾的场景中优化计划者，并且除了最终轨迹进行决策外，它还可以提供关键行为的预测。实验结果表明，我们的方法在轨迹产生和预测方面都可以实现有希望的表现。
Trajectory generation and prediction are two interwoven tasks that play important roles in planner evaluation and decision making for intelligent vehicles. Most existing methods focus on one of the two and are optimized to directly output the final generated/predicted trajectories, which only contain limited information for critical scenario augmentation and safe planning. In this work, we propose a novel behavior-aware Trajectory Autoencoder (TAE) that explicitly models drivers' behavior such as aggressiveness and intention in the latent space, using semi-supervised adversarial autoencoder and domain knowledge in transportation. Our model addresses trajectory generation and prediction in a unified architecture and benefits both tasks: the model can generate diverse, controllable and realistic trajectories to enhance planner optimization in safety-critical and long-tailed scenarios, and it can provide prediction of critical behavior in addition to the final trajectories for decision making. Experimental results demonstrate that our method achieves promising performance on both trajectory generation and prediction.
