Physics-informed Diffusion Mamba Transformer for Real-world Driving

TL;DR

This paper presents a novel diffusion transformer model for autonomous driving that effectively captures temporal dependencies and physical laws, improving trajectory prediction accuracy and safety.

Contribution

It introduces a diffusion Mamba Transformer architecture with integrated physical constraints, advancing multi-modal motion prediction in autonomous driving.

Findings

Outperforms state-of-the-art models in accuracy

Enhances physical plausibility of predictions

Improves robustness in diverse driving scenarios

Abstract

Autonomous driving systems demand trajectory planners that not only model the inherent uncertainty of future motions but also respect complex temporal dependencies and underlying physical laws. While diffusion-based generative models excel at capturing multi-modal distributions, they often fail to incorporate long-term sequential contexts and domain-specific physical priors. In this work, we bridge these gaps with two key innovations. First, we introduce a Diffusion Mamba Transformer architecture that embeds mamba and attention into the diffusion process, enabling more effective aggregation of sequential input contexts from sensor streams and past motion histories. Second, we design a Port-Hamiltonian Neural Network module that seamlessly integrates energy-based physical constraints into the diffusion model, thereby enhancing trajectory predictions with both consistency and interpretability. Extensive evaluations on standard autonomous driving benchmarks demonstrate that our unified framework significantly outperforms state-of-the-art baselines in predictive accuracy, physical plausibility, and robustness, thereby advancing safe and reliable motion planning.