Official repository for 3D vision research projects from the Media Computing Lab, Nankai University.
This repository collects code, data preparation scripts, evaluation tools, and project documentation for research on 3D vision, point cloud understanding, visual localization, and vision-language reasoning in 3D scenes. Each research project is maintained in its own folder with a dedicated README for installation, data preparation, training, inference, evaluation, and citation details.
| Folders | Publications | Venues |
|---|---|---|
vlm-loc |
VLM-Loc: Localization in Point Cloud Maps via Vision-Language Models | CVPR 2026 (Highlight) |
If this repository is helpful, please consider giving it a star. If this project benefits your work, please consider citing the corresponding paper.
@inproceedings{kang2026vlm,
title={VLM-Loc: Localization in Point Cloud Maps via Vision-Language Models},
author={Kang, Shuhao and Liao, Youqi and Wang, Peijie and Liao, Wenlong and Zhang, Qilin and Busam, Benjamin and Chen, Xieyuanli and Liu, Yun},
booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
pages={41365--41375},
year={2026}
}This repository is maintained by the Media Computing Lab, Nankai University. New 3D vision projects will be added as independent subdirectories.
VLM-Loc contact: kang-1-2-3 (kangshuhao1@163.com).