用图注意力与Mamba融合模型提升交通轨迹预测精度
GAMDTP: Dynamic Trajectory Prediction with Graph Attention Mamba Network
- 融合自注意力与Mamba-SSM，通过门控机制优化特征提取
- 在Argoverse数据集上实现当前最优轨迹预测性能
- 适合关注自动驾驶动态预测的算法研究者
准确预测交通参与者运动轨迹对自动驾驶系统的安全与稳定至关重要。本文提出GAMDTP，一种基于图注意力的动态轨迹预测网络。该方法在每一层图卷积中，通过门控机制融合自注意力与Mamba-SSM的结果，更高效精准地提取特征。GAMDTP编码高精地图（HD map）数据及车辆历史轨迹坐标，并解码输出最终预测结果。此外，针对现有方法多依赖两阶段框架（生成+优化）且动态融合历史预测的问题，设计评分机制，在生成与优化阶段评估预测质量以提升性能。在Argoverse数据集上的实验表明，GAMDTP达到当前最优水平，显著提升动态轨迹预测准确性。
原文摘要 · Abstract (English)
Accurate motion prediction of traffic agents is crucial for the safety and stability of autonomous driving systems. In this paper, we introduce GAMDTP, a novel graph attention-based network tailored for dynamic trajectory prediction. Specifically, we fuse the result of self attention and mamba-ssm through a gate mechanism, leveraging the strengths of both to extract features more efficiently and accurately, in each graph convolution layer. GAMDTP encodes the high-definition map(HD map) data and the agents' historical trajectory coordinates and decodes the network's output to generate the final prediction results. Additionally, recent approaches predominantly focus on dynamically fusing historical forecast results and rely on two-stage frameworks including proposal and refinement. To further enhance the performance of the two-stage frameworks we also design a scoring mechanism to evaluate the prediction quality during the proposal and refinement processes. Experiments on the Argoverse dataset demonstrates that GAMDTP achieves state-of-the-art performance, achieving superior accuracy in dynamic trajectory prediction.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。