HERMES: A Unified Self-Driving World Model for Simultaneous 3D Scene Understanding and Generation

TL;DR

HERMES is a unified model that combines 3D scene understanding and generation for autonomous driving, leveraging BEV representations and world queries to improve accuracy and contextual reasoning.

Contribution

It introduces a novel unified framework integrating scene understanding and generation using BEV and world queries, advancing autonomous driving scene modeling.

Findings

Reduces scene generation error by 32.4%.

Improves understanding metrics like CIDEr by 8.0%.

Validates effectiveness on nuScenes datasets.

Abstract

Driving World Models (DWMs) have become essential for autonomous driving by enabling future scene prediction. However, existing DWMs are limited to scene generation and fail to incorporate scene understanding, which involves interpreting and reasoning about the driving environment. In this paper, we present a unified Driving World Model named HERMES. We seamlessly integrate 3D scene understanding and future scene evolution (generation) through a unified framework in driving scenarios. Specifically, HERMES leverages a Bird's-Eye View (BEV) representation to consolidate multi-view spatial information while preserving geometric relationships and interactions. We also introduce world queries, which incorporate world knowledge into BEV features via causal attention in the Large Language Model, enabling contextual enrichment for understanding and generation tasks. We conduct comprehensive studies on nuScenes and OmniDrive-nuScenes datasets to validate the effectiveness of our method. HERMES achieves state-of-the-art performance, reducing generation error by 32.4% and improving understanding metrics such as CIDEr by 8.0%. The model and code will be publicly released at https://github.com/LMD0311/HERMES.