ReconDrive: Fast Feed-Forward 4D Gaussian Splatting for Autonomous Driving Scene Reconstruction

TL;DR

ReconDrive introduces a fast, feed-forward 4D Gaussian Splatting framework for autonomous driving scene reconstruction, combining high fidelity and scalability by leveraging a foundation model and novel adaptations.

Contribution

It extends the 3D foundation model VGGT with hybrid Gaussian prediction heads and a static-dynamic composition strategy for dynamic scene modeling.

Findings

Outperforms existing feed-forward methods in reconstruction and view synthesis

Achieves near per-scene optimization quality with much faster speed

Demonstrates effectiveness on nuScenes dataset for realistic driving simulation

Abstract

High-fidelity visual reconstruction and novel-view synthesis are essential for realistic closed-loop evaluation in autonomous driving. While 4D Gaussian Splatting (4DGS) offers a promising balance of accuracy and efficiency, existing per-scene optimization methods require costly iterative refinement, rendering them unscalable for extensive urban environments. Conversely, current feed-forward approaches often suffer from degraded photometric quality. To address these limitations, we propose ReconDrive, a feed-forward framework that leverages and extends the 3D foundation model VGGT for rapid, high-fidelity 4DGS generation. Our architecture introduces two core adaptations to tailor the foundation model to dynamic driving scenes: (1) Hybrid Gaussian Prediction Heads, which decouple the regression of spatial coordinates and appearance attributes to overcome the photometric deficiencies inherent in generalized foundation features; and (2) a Static-Dynamic 4D Composition strategy that explicitly captures temporal motion via velocity modeling to represent complex dynamic environments. Benchmarked on nuScenes, ReconDrive significantly outperforms existing feed-forward baselines in reconstruction, novel-view synthesis, and 3D perception. It achieves performance competitive with per-scene optimization while being orders of magnitude faster, providing a scalable and practical solution for realistic driving simulation.