Driv3R: Learning Dense 4D Reconstruction for Autonomous Driving

TL;DR

Driv3R is a novel framework for real-time dense 4D scene reconstruction in autonomous driving, leveraging multi-view images, dynamic object detection, and a memory pool for improved accuracy and speed.

Contribution

It introduces a direct per-frame point map regression method with a memory pool and 4D flow predictor, enabling fast, streaming 4D reconstruction without global alignment.

Findings

Achieves 15x faster inference speed than previous methods

Outperforms existing frameworks in 4D dynamic scene reconstruction

Effectively reconstructs dynamic scenes in real-time

Abstract

Realtime 4D reconstruction for dynamic scenes remains a crucial challenge for autonomous driving perception. Most existing methods rely on depth estimation through self-supervision or multi-modality sensor fusion. In this paper, we propose Driv3R, a DUSt3R-based framework that directly regresses per-frame point maps from multi-view image sequences. To achieve streaming dense reconstruction, we maintain a memory pool to reason both spatial relationships across sensors and dynamic temporal contexts to enhance multi-view 3D consistency and temporal integration. Furthermore, we employ a 4D flow predictor to identify moving objects within the scene to direct our network focus more on reconstructing these dynamic regions. Finally, we align all per-frame pointmaps consistently to the world coordinate system in an optimization-free manner. We conduct extensive experiments on the large-scale nuScenes dataset to evaluate the effectiveness of our method. Driv3R outperforms previous frameworks in 4D dynamic scene reconstruction, achieving 15x faster inference speed compared to methods requiring global alignment. Code: https://github.com/Barrybarry-Smith/Driv3R.