Multiagent Trajectory Prediction With Difficulty-Guided Feature Enhancement Network
Guipeng Xin, Duanfeng Chu, LU Li-ping, Zejian Deng, Yuang Lu, Xigang Wu
Abstract
Trajectory prediction is crucial for autonomous driving, as it aims to forecast the future movements of traffic participants. Traditional methods usually perform holistic inference on the trajectories of agents, neglecting differences in prediction difficulty among agents. This letter proposes a novel Difficulty-Guided Feature Enhancement Network (DGFNet), which leverages the prediction difficulty differences among agents for multi-agent trajectory prediction. Firstly, we employ Spatio-temporal Feature Extraction to capture rich spatio-temporal features. Secondly, a Difficulty-Guided Decoder controls the flow of future trajectories into subsequent modules, obtaining reliable future trajectories. Then, feature interaction and fusion are performed through the Future Feature Interaction module. Finally, the fused actor features are fed into the Final Decoder to generate the predicted trajectory distributions for multiple participants. Experimental results demonstrate that our model achieves SOTA performance on the Argoverse 1&2 motion forecasting benchmarks. Ablation studies further validate the effectiveness of each module. Moreover, compared to the SOTA methods, our method balances trajectory prediction accuracy and real-time inference speed.