GAMDTP: Dynamic Trajectory Prediction with Graph Attention Mamba Network
This work addresses motion prediction for autonomous driving systems, representing an incremental improvement with novel method integration.
The paper tackles dynamic trajectory prediction for autonomous driving by introducing GAMDTP, a graph attention-based network that fuses self-attention and mamba-ssm with a gate mechanism and a scoring mechanism for two-stage frameworks, achieving state-of-the-art performance on the Argoverse dataset.
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.