360DVO: Deep Visual Odometry for Monocular 360-Degree Camera

The Hong Kong University of Science and Technology

IEEE Robotics and Automation Letters (RA-L), 2026.

Xiaopeng Guo, Yinzhe Xu, Huajian Huang, Sai-Kit Yeung

Getting Started

Installation

1. Clone 360DVO.

git clone https://github.com/chris1004336379/360DVO.git
cd 360DVO

2. Create the environment.

conda env create -f environment.yml
conda activate 360dvo

3. Install the 360DVO package

wget https://gitlab.com/libeigen/eigen/-/archive/3.4.0/eigen-3.4.0.zip
unzip eigen-3.4.0.zip -d thirdparty

# install
pip install .

4. Download weights

Demo

360DVO can be run on any 360 video or image directory with a single command. We use Viser to visualize the reconstructions in real-time.

python demo.py \
    --imagedir=<path to image directory or video> \
    --viz # enable visualization
    --plot # save trajectory plot
    --save_ply # save point cloud as a .ply file
    --save_trajectory # save the predicted trajectory as .txt in TUM format

Training

comming soon

Acknowledgements

Our code is based on DPVO, SphereNet, Viser. We thank the authors for their excellent work!

Citation

If you find our work useful, please cite:

@ARTICLE{11358682,
  author={Guo, Xiaopeng and Xu, Yinzhe and Huang, Huajian and Yeung, Sai-Kit},
  journal={IEEE Robotics and Automation Letters}, 
  title={360DVO: Deep Visual Odometry for Monocular 360-Degree Camera}, 
  year={2026},
  volume={11},
  number={3},
  pages={3079-3086},
  keywords={Feature extraction;Cameras;Nonlinear distortion;Convolution;Kernel;Bundle adjustment;Visual odometry;Accuracy;Benchmark testing;Robustness;Visual odometry;omnidirectional vision},
  doi={10.1109/LRA.2026.3655280}}