Current browse context:
cs.CV
Change to browse by:
References & Citations
Computer Science > Computer Vision and Pattern Recognition
Title: EvolveGraph: Multi-Agent Trajectory Prediction with Dynamic Relational Reasoning
(Submitted on 31 Mar 2020 (v1), last revised 22 Oct 2020 (this version, v4))
Abstract: Multi-agent interacting systems are prevalent in the world, from pure physical systems to complicated social dynamic systems. In many applications, effective understanding of the situation and accurate trajectory prediction of interactive agents play a significant role in downstream tasks, such as decision making and planning. In this paper, we propose a generic trajectory forecasting framework (named EvolveGraph) with explicit relational structure recognition and prediction via latent interaction graphs among multiple heterogeneous, interactive agents. Considering the uncertainty of future behaviors, the model is designed to provide multi-modal prediction hypotheses. Since the underlying interactions may evolve even with abrupt changes, and different modalities of evolution may lead to different outcomes, we address the necessity of dynamic relational reasoning and adaptively evolving the interaction graphs. We also introduce a double-stage training pipeline which not only improves training efficiency and accelerates convergence, but also enhances model performance. The proposed framework is evaluated on both synthetic physics simulations and multiple real-world benchmark datasets in various areas. The experimental results illustrate that our approach achieves state-of-the-art performance in terms of prediction accuracy.
Submission history
From: Jiachen Li [view email][v1] Tue, 31 Mar 2020 02:49:23 GMT (2833kb,D)
[v2] Mon, 13 Apr 2020 07:28:20 GMT (2803kb,D)
[v3] Thu, 1 Oct 2020 17:42:10 GMT (6048kb,D)
[v4] Thu, 22 Oct 2020 16:02:44 GMT (6050kb,D)
Link back to: arXiv, form interface, contact.