DDP-WM: Disentangled Dynamics Prediction for Efficient World Models

TL;DR

DDP-WM introduces a disentangled dynamics prediction approach that significantly improves efficiency and accuracy of world models in robotic tasks by decomposing scene dynamics into primary and secondary components.

Contribution

The paper proposes a novel architecture for world models that decomposes latent state evolution into primary and secondary dynamics, enhancing efficiency and performance.

Findings

Achieves approximately 9x inference speedup on Push-T task.

Improves MPC success rate from 90% to 98%.

Demonstrates effectiveness across diverse robotic tasks.

Abstract

World models are essential for autonomous robotic planning. However, the substantial computational overhead of existing dense Transformerbased models significantly hinders real-time deployment. To address this efficiency-performance bottleneck, we introduce DDP-WM, a novel world model centered on the principle of Disentangled Dynamics Prediction (DDP). We hypothesize that latent state evolution in observed scenes is heterogeneous and can be decomposed into sparse primary dynamics driven by physical interactions and secondary context-driven background updates. DDP-WM realizes this decomposition through an architecture that integrates efficient historical processing with dynamic localization to isolate primary dynamics. By employing a crossattention mechanism for background updates, the framework optimizes resource allocation and provides a smooth optimization landscape for planners. Extensive experiments demonstrate that DDP-WM achieves significant efficiency and performance across diverse tasks, including navigation, precise tabletop manipulation, and complex deformable or multi-body interactions. Specifically, on the challenging Push-T task, DDP-WM achieves an approximately 9 times inference speedup and improves the MPC success rate from 90% to98% compared to state-of-the-art dense models. The results establish a promising path for developing efficient, high-fidelity world models. Codes is available at https://hcplab-sysu.github.io/DDP-WM/.