ReCamDriving: LiDAR-Free Camera-Controlled Novel Trajectory Video Generation
This addresses the challenge of camera-controllable video generation for applications like autonomous driving or simulation, though it appears incremental by building on existing 3DGS and video generation methods.
The paper tackles the problem of generating novel-trajectory videos from monocular inputs by proposing ReCamDriving, a vision-based framework that uses 3DGS renderings for geometric guidance and a two-stage training approach, achieving state-of-the-art camera controllability and structural consistency.
We propose ReCamDriving, a purely vision-based, camera-controlled novel-trajectory video generation framework. While repair-based methods fail to restore complex artifacts and LiDAR-based approaches rely on sparse and incomplete cues, ReCamDriving leverages dense and scene-complete 3DGS renderings for explicit geometric guidance, achieving precise camera-controllable generation. To mitigate overfitting to restoration behaviors when conditioned on 3DGS renderings, ReCamDriving adopts a two-stage training paradigm: the first stage uses camera poses for coarse control, while the second stage incorporates 3DGS renderings for fine-grained viewpoint and geometric guidance. Furthermore, we present a 3DGS-based cross-trajectory data curation strategy to eliminate the train-test gap in camera transformation patterns, enabling scalable multi-trajectory supervision from monocular videos. Based on this strategy, we construct the ParaDrive dataset, containing over 110K parallel-trajectory video pairs. Extensive experiments demonstrate that ReCamDriving achieves state-of-the-art camera controllability and structural consistency.