[AAAI'26] MoBGS: Motion Deblurring Dynamic 3D Gaussian Splatting for Blurry Monocular Video

Minh-Quan Viet Bui^*, Jongmin Park^*, Juan Luis Gonzalez Bello, Jaeho Moon, Jihyong Oh^†, Munchurl Kim^†

*Co-first authors (equal contribution), †Co-corresponding authors

⚙️ Environmental Setups

Please refer to the environment installation of SplineGS, and install gsplat, simple-knn as

pip install git+https://github.com/nerfstudio-project/gsplat.git@v1.4.0
pip install -e submodules/simple-knn

📁 Data Preparations

We follow the evaluation setup from DyBluRF. Download our preprocessed dataset here and arrange them as follows:

MoBGS/data/stereo
    ├── basketball
    │   
    │   
    │   
    │   
    ├── ...
    └── street

🚀 Get Started

Training

python train.py -s data/stereo/seesaw/dense/ --port 6969 --expname "seesaw" --configs arguments/stereo/seesaw.py

Metrics Evaluation

python eval.py -s data/stereo/seesaw/dense/ --port 6018 --expname "seesaw" --configs arguments/stereo/seesaw.py --checkpoint output/seesaw/point_cloud/iteration_10000

python metrics.py --datadir data/stereo/seesaw/dense/ --scene_name seesaw --output_dir output

Acknowledgments

• This work was supported by Institute of Information and communications Technology Planning and Evaluation (IITP) grant funded by the Korean Government [Ministry of Science and ICT (Information and Communications Technology)] (Project Number: RS-2022-00144444, Project Title: Deep Learning Based Visual Representational Learning and Rendering of Static and Dynamic Scenes, 100%).

⭐ Citing MoBGS

If you find our repository useful, please consider giving it a star ⭐ and citing our research papers in your work:

@InProceedings{bui2025mobgs,
    author    = {Bui, Minh-Quan Viet and Park, Jongmin and Bello, Juan Luis Gonzalez and Moon, Jaeho and Oh, Jihyong and Kim, Munchurl},
    title     = {MoBGS: Motion Deblurring Dynamic 3D Gaussian Splatting for Blurry Monocular Video},
    booktitle = {AAAI},
    year      = {2026},
}

📈 Star History