Multi-agent Trajectory Prediction with Fuzzy Query Attention
arXiv:2010.15891
Abstract
Trajectory prediction for scenes with multiple agents and entities is a challenging problem in numerous domains such as traffic prediction, pedestrian tracking and path planning. We present a general architecture to address this challenge which models the crucial inductive biases of motion, namely, inertia, relative motion, intents and interactions. Specifically, we propose a relational model to flexibly model interactions between agents in diverse environments. Since it is well-known that human decision making is fuzzy by nature, at the core of our model lies a novel attention mechanism which models interactions by making continuous-valued (fuzzy) decisions and learning the corresponding responses. Our architecture demonstrates significant performance gains over existing state-of-the-art predictive models in diverse domains such as human crowd trajectories, US freeway traffic, NBA sports data and physics datasets. We also present ablations and augmentations to understand the decision-making process and the source of gains in our model.
NeurIPS 2020 Camera-ready version. Code: https://github.com/nitinkamra1992/FQA
References in corpus (10)
- Adam: A Method for Stochastic Optimization
- Social Force Model for Pedestrian Dynamics
- Inductive Representation Learning on Large Graphs
- Learning to Simulate Complex Physics with Graph Networks
- A Compositional Object-Based Approach to Learning Physical Dynamics
- Human Trajectory Prediction using Spatially aware Deep Attention Models
- Generating Long-term Trajectories Using Deep Hierarchical Networks
- An Evaluation of Trajectory Prediction Approaches and Notes on the TrajNet Benchmark
- Relational Forward Models for Multi-Agent Learning
- Stochastic Prediction of Multi-Agent Interactions from Partial Observations