TL;DR: MuGS is a feedforward 3D scene reconstruction framework for multiple camera baseline settings.

We present Multi-Baseline Gaussian Splatting (MuGS), a generalized feed-forward approach for novel view synthesis that effectively handles diverse baseline settings, including sparse input views with both small and large baselines. Specifically, we integrate features from Multi-View Stereo (MVS) and Monocular Depth Estimation (MDE) to enhance feature representations for generalizable reconstruction. Next, We propose a projection-and-sampling mechanism for deep depth fusion, which constructs a fine probability volume to guide the regression of the feature map. Furthermore, We introduce a reference-view loss to improve geometry and optimization efficiency. We leverage 3D Gaussian representations to accelerate training and inference time while enhancing rendering quality. MuGS achieves state-of-the-art performance across multiple baseline settings and diverse scenarios ranging from simple objects (DTU) to complex indoor and outdoor scenes (RealEstate10K). We also demonstrate promising zero-shot performance on the LLFF and Mip-NeRF 360 datasets.

If you find our work useful for your research, please consider citing our paper:

@InProceedings{Lou_2025_ICCV,
    author    = {Lou, Yaopeng and Shen, Liao and Liu, Tianqi and Li, Jiaqi and Huang, Zihao and Sun, Huiqiang and Cao, Zhiguo},
    title     = {MuGS: Multi-Baseline Generalizable Gaussian Splatting Reconstruction},
    booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)},
    month     = {October},
    year      = {2025},
    pages     = {25583-25593}
}

This work is built on many amazing open-source projects shared by MuRF, DepthAnythingV2 and 3DGS. Thanks all the authors for their excellent contributions!