Another Vertical View: A Hierarchical Network for Heterogeneous Trajectory Prediction via Spectrums

TL;DR

This paper introduces a hierarchical network that uses spectrums and Haar transforms to predict heterogeneous trajectories, effectively modeling complex dimension-wise interactions across various trajectory forms.

Contribution

It extends trajectory prediction to heterogeneous data by incorporating spectrum analysis and a bilinear fusion approach, addressing the limitations of prior methods focused on specific trajectory types.

Findings

Outperforms state-of-the-art methods on multiple datasets

Effective modeling of heterogeneous trajectories including coordinates, bounding boxes, and skeletons

Captures complex dimension-wise interactions using spectrum-based hierarchical modeling

Abstract

With the fast development of AI-related techniques, the applications of trajectory prediction are no longer limited to easier scenes and trajectories. More and more trajectories with different forms, such as coordinates, bounding boxes, and even high-dimensional human skeletons, need to be analyzed and forecasted. Among these heterogeneous trajectories, interactions between different elements within a frame of trajectory, which we call ``Dimension-wise Interactions'', would be more complex and challenging. However, most previous approaches focus mainly on a specific form of trajectories, and potential dimension-wise interactions are less concerned. In this work, we expand the trajectory prediction task by introducing the trajectory dimensionality $M$, thus extending its application scenarios to heterogeneous trajectories. We first introduce the Haar transform as an alternative to Fourier transform to better capture the time-frequency properties of each trajectory-dimension. Then, we adopt the bilinear structure to model and fuse two factors simultaneously, including the time-frequency response and the dimension-wise interaction, to forecast heterogeneous trajectories via trajectory spectrums hierarchically in a generic way. Experiments show that the proposed model outperforms most state-of-the-art methods on ETH-UCY, SDD, nuScenes, and Human3.6M with heterogeneous trajectories, including 2D coordinates, 2D/3D bounding boxes, and 3D human skeletons.