Conditional Neural Relational Inference for Interacting Systems

TL;DR

This paper introduces a conditional neural relational inference model that can generate dynamics of interacting systems from a vectorial description without needing partial trajectory data, demonstrated on human gait modeling.

Contribution

It presents a novel conditional generative model for dynamical systems that does not require trajectory input during generation, unlike previous methods.

Findings

Successfully models human gait dynamics, including pathological cases.

Outperforms existing models in trajectory generation tasks.

Provides a flexible framework for interacting systems with shared physical laws.

Abstract

In this work, we want to learn to model the dynamics of similar yet distinct groups of interacting objects. These groups follow some common physical laws that exhibit specificities that are captured through some vectorial description. We develop a model that allows us to do conditional generation from any such group given its vectorial description. Unlike previous work on learning dynamical systems that can only do trajectory completion and require a part of the trajectory dynamics to be provided as input in generation time, we do generation using only the conditioning vector with no access to generation time's trajectories. We evaluate our model in the setting of modeling human gait and, in particular pathological human gait.