TY - CHAP

T1 - Semi-Supervised Generative Models for Multi-Agent Trajectories

AU - Brefeld, Ulf

AU - Fassmeyer, Dennis

AU - Fassmeyer, Pascal

N1 - Conference code: 36

PY - 2022

Y1 - 2022

N2 - Analyzing the spatiotemporal behavior of multiple agents is of great interest to many communities. Existing probabilistic models in this realm are formalized either in an unsupervised framework, where the latent space is described by discrete or continuous variables, or in a supervised framework, where weakly preserved labels add explicit information to continuous latent representations. To overcome inherent limitations, we propose a novel objective function for processing multi-agent trajectories based on semi-supervised variational autoencoders, where equivariance and interaction of agents are captured via customized graph networks. The resulting architecture disentangles discrete and continuous latent effects and provides a natural solution for injecting expensive domain knowledge into interactive sequential systems. Empirically, our model not only outperforms various state-of-the-art baselines in trajectory forecasting, but also learns to effectively leverage unsupervised multi-agent sequences for classification tasks on interactive real-world sports datasets.

AB - Analyzing the spatiotemporal behavior of multiple agents is of great interest to many communities. Existing probabilistic models in this realm are formalized either in an unsupervised framework, where the latent space is described by discrete or continuous variables, or in a supervised framework, where weakly preserved labels add explicit information to continuous latent representations. To overcome inherent limitations, we propose a novel objective function for processing multi-agent trajectories based on semi-supervised variational autoencoders, where equivariance and interaction of agents are captured via customized graph networks. The resulting architecture disentangles discrete and continuous latent effects and provides a natural solution for injecting expensive domain knowledge into interactive sequential systems. Empirically, our model not only outperforms various state-of-the-art baselines in trajectory forecasting, but also learns to effectively leverage unsupervised multi-agent sequences for classification tasks on interactive real-world sports datasets.

KW - Informatics

KW - Business informatics

UR - http://www.scopus.com/inward/record.url?scp=85163202847&partnerID=8YFLogxK

UR - https://www.proceedings.com/content/068/068431webtoc.pdf

M3 - Article in conference proceedings

AN - SCOPUS:85163202847

SN - 978-1-7138-7108-8

VL - 48

T3 - Advances in Neural Information Processing Systems

SP - 37267

EP - 37281

BT - Advances in Neural Information Processing Systems 35

A2 - Koyejo, S.

A2 - Mohamed, S.

A2 - Agarwal, A.

A2 - Belgrave, D.

A2 - Cho, K.

A2 - Oh, A.

PB - Curran Associates

CY - Red Hook

T2 - 36th Conference on Neural Information Processing Systems - NeurIPS 2022

Y2 - 28 November 2022 through 9 December 2022

ER -