Object-centric Denoising Diffusion Models for Physical Reasoning

TL;DR

This paper introduces an object-centric denoising diffusion model for physical reasoning that handles multiple conditions, is permutation and translation equivariant, and adapts to varying object counts and trajectory lengths.

Contribution

It presents a novel diffusion-based architecture for physical reasoning that overcomes limitations of autoregressive models by enabling flexible conditioning on multiple time steps and objects.

Findings

Effective in solving multi-condition physical reasoning tasks

Handles varying object numbers during inference

Maintains performance across different trajectory lengths

Abstract

Reasoning about the trajectories of multiple, interacting objects is integral to physical reasoning tasks in machine learning. This involves conditions imposed on the objects at different time steps, for instance initial states or desired goal states. Existing approaches in physical reasoning generally rely on autoregressive modeling, which can only be conditioned on initial states, but not on later states. In fields such as planning for reinforcement learning, similar challenges are being addressed with denoising diffusion models. In this work, we propose an object-centric denoising diffusion model architecture for physical reasoning that is translation equivariant over time, permutation equivariant over objects, and can be conditioned on arbitrary time steps for arbitrary objects. We demonstrate how this model can solve tasks with multiple conditions and examine its performance when changing object numbers and trajectory lengths during inference.