车辆轨迹预测：基于多EKF轨迹候选的神经融合方法
Vehicle Trajectory Prediction via Neural Fusion of Multiple EKF-Based Trajectory Candidates
浏览论文内容
中文总结 AI 辅助
本研究提出一种融合Trajectron++与多EKF轨迹候选的框架，在nuScenes上使平均和最终位移误差分别降低13.7%和14.6%，提升物理可行性。
中文摘要 AI 辅助
在自动驾驶中，预测周围车辆的未来轨迹对于碰撞风险评估和安全自车路径规划至关重要。传统的基于神经网络的轨迹预测器通常通过利用智能体历史信息、动态场景图和语义地图来实现强大的预测性能。然而，在特定的运动模式（如加速、减速和转弯）下，这些预测器可能无法反映物理上可行的轨迹。为解决这一问题，本研究提出了一种框架，在后期阶段将基于神经网络的轨迹预测器Trajectron++的输出与基于扩展卡尔曼滤波（EKF）的多个轨迹候选进行融合。在nuScenes数据集上，所提出的方法在不修改基线架构的情况下，将Trajectron++机器人基线的平均位移误差和最终位移误差分别降低了13.7%和14.6%。这些结果表明，基于EKF的轨迹候选可以通过学习融合有效补充神经轨迹预测。
英文摘要
Predicting the future trajectories of surrounding vehicles in autonomous driving is important for collision risk assessment and safe ego-vehicle path planning. Conventional neural network-based trajectory predictors typically achieve strong prediction performance by exploiting agent history, dynamic scene graphs, and semantic maps. However, in specific motion regimes such as acceleration, deceleration, and turning, these predictors may fail to reflect physically feasible trajectories. To address this issue, this study proposes a framework that fuses the output of Trajectron++, a neural network-based trajectory predictor, with extended Kalman filter (EKF)-based multiple trajectory candidates at a late stage. On the nuScenes dataset, the proposed method reduces the average displacement error and final displacement error of the Trajectron++ robot baseline by 13.7% and 14.6%, respectively, without modifying the baseline architecture. These results indicate that EKF-based trajectory candidates can effectively complement neural trajectory prediction through learned fusion.
发表机构
- Korea Advanced Institute of Science and Technology (KAIST)（韩国科学技术院（KAIST））
机构由 AI 辅助整理，请以论文原文为准。
