Conditional Neural Relational Inference for Interacting Systems
This work addresses the challenge of generating dynamics for specific groups in interacting systems, such as human gait analysis, but appears incremental as it builds on prior neural relational inference methods by adding conditional capabilities.
The paper tackles the problem of modeling the dynamics of distinct groups of interacting objects by developing a model for conditional generation using only a vectorial description, without requiring trajectory inputs at generation time, and demonstrates its application in modeling human gait, including pathological cases.
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.