Studying the Effect of Explicit Interaction Representations on Learning Scene-level Distributions of Human Trajectories

2025-11-06

AI Summary

This paper addresses the problem of modeling joint trajectory distributions for multiple agents in autonomous driving scenarios, with a focus on effectively representing inter-agent interactions to improve scene-level prediction accuracy. Methodologically, it systematically compares explicit interaction representations—such as spatiotemporal relational modeling grounded in traffic rules—against implicit, neural-network-based interaction learning, all within a unified architectural framework, and quantitatively evaluates their impact on joint distribution learning. Experimental results demonstrate that explicit interaction modeling—particularly when incorporating domain priors (e.g., right-of-way rules at intersections)—yields substantial gains in both prediction accuracy and robustness, challenging the prevailing assumption that implicit learning is inherently superior. The primary contribution is the first empirical demonstration of the superiority of explicit interaction modeling for scene-level trajectory forecasting, thereby establishing a new paradigm for interpretable and reliable autonomous driving decision-making.

Technology Category

• Multiagent Systems: Adversarial Agents
• Planning, Routing, and Scheduling: Model-Based Reasoning
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Effectively capturing the joint distribution of all agents in a scene is relevant for predicting the true evolution of the scene and in turn providing more accurate information to the decision processes of autonomous vehicles. While new models have been developed for this purpose in recent years, it remains unclear how to best represent the joint distributions particularly from the perspective of the interactions between agents. Thus far there is no clear consensus on how best to represent interactions between agents; whether they should be learned implicitly from data by neural networks, or explicitly modeled using the spatial and temporal relations that are more grounded in human decision-making. This paper aims to study various means of describing interactions within the same network structure and their effect on the final learned joint distributions. Our findings show that more often than not, simply allowing a network to establish interactive connections between agents based on data has a detrimental effect on performance. Instead, having well defined interactions (such as which agent of an agent pair passes first at an intersection) can often bring about a clear boost in performance.

Problem

• Studying how interaction representations affect learning human trajectory distributions
• Comparing implicit vs explicit modeling of agent interactions for scene prediction
• Evaluating interaction representation effects on autonomous vehicle decision accuracy

Innovation

• Explicit interaction modeling improves trajectory distribution learning
• Data-driven implicit interactions often harm predictive performance
• Well-defined agent interactions boost autonomous vehicle decision accuracy