DrivingGaussian: Composite Gaussian Splatting for Surrounding Dynamic Autonomous Driving Scenes
Authors:
Preprint
Abstract
We present DrivingGaussian, an efficient and effective framework for surrounding dynamic autonomous driving scenes. For complex scenes with moving objects, we first sequentially and progressively model the static background of the entire scene with incremental static 3D Gaussians. We then leverage a composite dynamic Gaussian graph to handle multiple moving objects, individually reconstructing each object and restoring their accurate positions and occlusion relationships within the scene. We further use a LiDAR prior for Gaussian Splatting to reconstruct scenes with greater details and maintain panoramic consistency. DrivingGaussian outperforms existing methods in driving scene reconstruction and enables photorealistic surround-view synthesis with high-fidelity and multi-camera consistency. Our project page is at: https://github.com/VDIGPKU/DrivingGaussian.Keywords
All Related Versions
This publication has 34 references indexed in Scilit:
- Dynamic 3D Gaussians: Tracking by Persistent Dynamic View Synthesis2024
- MARS: An Instance-Aware, Modular and Realistic Simulator for Autonomous Driving2024
- BEVFormer: Learning Bird’s-Eye-View Representation from Multi-camera Images via Spatiotemporal Transformers2022
- Plenoxels: Radiance Fields without Neural Networks2022
- Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields2022
- Point-NeRF: Point-based Neural Radiance Fields2022
- Block-NeRF: Scalable Large Scene Neural View Synthesis2022
- D-NeRF: Neural Radiance Fields for Dynamic Scenes2021
- Neural Scene Graphs for Dynamic Scenes2021
- nuScenes: A Multimodal Dataset for Autonomous Driving2020