EVolSplat: Efficient Volume-based Gaussian Splatting for Urban View Synthesis

CVPR 2025

1Zhejiang University, 2Huawei Noah's Ark Lab, 3University of Tübingen, 4Tübingen AI Center

Overview

Teaser image demonstrating Marigold depth estimation.

This paper presents EVolSplat, a method for efficient urban scene reconstruction in a feed-forward manner. Unlike previous pixel-aligned 3DGS frameworks, we use geometric priors to construct a global volume and predict a standalone 3D representation, achieving state-of-the-art performance across several street-view datasets and enabling real-time rendering, making it well-suited for urban scenes.

Video

Comparison with Feed-Forward Methods

How it works

EVolSplat learns to predict 3D Gaussians of urban scenes in a feed-forward manner. Given a set of posed images $\{I_n\}_{i=1}^N$, we first leverage off-the-shelf metric depth estimators to provide depth estimations $\{D_n\}_{n=1}^N$. The depth maps are unprojected and accumulated into a global point cloud $P$, which is fed into a sparse 3D CNN for extracting a feature volume $F$. We leverage the 3D context of $F$ to predict the geometry attributes of 3D Gaussians. Furthermore, we project the 3D Gaussians to the nearest reference views to retrieve 2D context, including color window $\{c_k\}_{k=1}^K$ and visibility maps $\{v_k\}_{k=1}^K$ to decode their color. To model far regions, we propose a generalizable hemisphere Gaussian model, where the geometry is fixed and the color is predicted in a similar manner as the foreground volume.

Marigold training scheme

Results Gallery

Feed-forward Reconstruction on KITTI-360

Zeroshot inference on Waymo (trained on KITTI-360)

BibTeX

@InProceedings{Miao_2025_CVPR,
      author    = {Sheng Miao, Jiaxin Huang, Dongfeng Bai, Xu Yan, Hongyu Zhou, Yue Wang, Bingbing Liu, Andreas Geiger and Yiyi Liao},
      title     = {EVolSplat: Efficient Volume-based Gaussian Splatting for Urban View Synthesis},
      booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
      month     = {June},
      year      = {2025}
  }