Self-Growing Spatial Graph Network for Context-Aware Pedestrian Trajectory Prediction
This work provides an incremental improvement for pedestrian trajectory prediction, specifically for adapting graph structures in unknown environments for online systems.
This paper addresses the challenge of adapting graph structures in unknown environments for online pedestrian trajectory prediction. The proposed STR-GGRNN model uses online Nonnegative Matrix Factorization (NMF) to adaptively recommend neighborhoods based on contextual scene features and pedestrian visual cues, achieving 12 cm ADE and ~15 cm FDE on the ETH-UCY dataset.
Pedestrian trajectory prediction is an active research area with recent works undertaken to embed accurate models of pedestrians social interactions and their contextual compliance into dynamic spatial graphs. However, existing works rely on spatial assumptions about the scene and dynamics, which entails a significant challenge to adapt the graph structure in unknown environments for an online system. In addition, there is a lack of assessment approach for the relational modeling impact on prediction performance. To fill this gap, we propose Social Trajectory Recommender-Gated Graph Recurrent Neighborhood Network, (STR-GGRNN), which uses data-driven adaptive online neighborhood recommendation based on the contextual scene features and pedestrian visual cues. The neighborhood recommendation is achieved by online Nonnegative Matrix Factorization (NMF) to construct the graph adjacency matrices for predicting the pedestrians' trajectories. Experiments based on widely-used datasets show that our method outperforms the state-of-the-art. Our best performing model achieves 12 cm ADE and $\sim$15 cm FDE on ETH-UCY dataset. The proposed method takes only 0.49 seconds when sampling a total of 20K future trajectories per frame.