SGBA: semantic gaussian mixture model-based LiDAR bundle adjustment
LiDAR bundle adjustment (BA) is an effective approach to reduce the drifts in pose estimation from the front-end. Existing works on LiDAR BA usually rely on predefined geometric features for landmark representation. This reliance restricts generalizability, as the system will inevitably deteriorate...
| Main Authors: | , , , , , |
|---|---|
| Other Authors: | |
| Format: | Journal Article |
| Language: | English |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/182120 |
| _version_ | 1826112123972878336 |
|---|---|
| author | Ji, Xingyu Yuan, Shenghai Li, Jianping Yin, Pengyu Cao, Haozhi Xie, Lihua |
| author2 | School of Electrical and Electronic Engineering |
| author_facet | School of Electrical and Electronic Engineering Ji, Xingyu Yuan, Shenghai Li, Jianping Yin, Pengyu Cao, Haozhi Xie, Lihua |
| author_sort | Ji, Xingyu |
| collection | NTU |
| description | LiDAR bundle adjustment (BA) is an effective approach to reduce the drifts in pose estimation from the front-end. Existing works on LiDAR BA usually rely on predefined geometric features for landmark representation. This reliance restricts generalizability, as the system will inevitably deteriorate in environments where these specific features are absent. To address this issue, we propose SGBA, a LiDAR BA scheme that models the environment as a semantic Gaussian mixture model (GMM) without predefined feature types. This approach encodes both geometric and semantic information, offering a comprehensive and general representation adaptable to various environments. Additionally, to limit computational complexity while ensuring generalizability, we propose an adaptive semantic selection framework that selects the most informative semantic clusters for optimization by evaluating the condition number of the cost function. Lastly, we introduce a probabilistic feature association scheme that considers the entire probability density of assignments, which can manage uncertainties in measurement and initial pose estimation. We have conducted various experiments and the results demonstrate that SGBA can achieve accurate and robust pose refinement even in challenging scenarios with low-quality initial pose estimation and limited geometric features. We plan to open source the work for the benefit of the community https://github.com/Ji1Xingyu/SGBA. |
| first_indexed | 2025-03-09T10:30:31Z |
| format | Journal Article |
| id | ntu-10356/182120 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2025-03-09T10:30:31Z |
| publishDate | 2025 |
| record_format | dspace |
| spelling | ntu-10356/1821202025-01-10T15:43:56Z SGBA: semantic gaussian mixture model-based LiDAR bundle adjustment Ji, Xingyu Yuan, Shenghai Li, Jianping Yin, Pengyu Cao, Haozhi Xie, Lihua School of Electrical and Electronic Engineering Centre for Advanced Robotics Technology Innovation (CARTIN) Computer and Information Science Localization Mapping Bundle adjustment LiDAR bundle adjustment (BA) is an effective approach to reduce the drifts in pose estimation from the front-end. Existing works on LiDAR BA usually rely on predefined geometric features for landmark representation. This reliance restricts generalizability, as the system will inevitably deteriorate in environments where these specific features are absent. To address this issue, we propose SGBA, a LiDAR BA scheme that models the environment as a semantic Gaussian mixture model (GMM) without predefined feature types. This approach encodes both geometric and semantic information, offering a comprehensive and general representation adaptable to various environments. Additionally, to limit computational complexity while ensuring generalizability, we propose an adaptive semantic selection framework that selects the most informative semantic clusters for optimization by evaluating the condition number of the cost function. Lastly, we introduce a probabilistic feature association scheme that considers the entire probability density of assignments, which can manage uncertainties in measurement and initial pose estimation. We have conducted various experiments and the results demonstrate that SGBA can achieve accurate and robust pose refinement even in challenging scenarios with low-quality initial pose estimation and limited geometric features. We plan to open source the work for the benefit of the community https://github.com/Ji1Xingyu/SGBA. National Research Foundation (NRF) Submitted/Accepted version This work was supported by National Research Foundation, Singapore, under its Medium-Sized Center for Advanced Robotics Technology Innovation. 2025-01-09T05:30:59Z 2025-01-09T05:30:59Z 2024 Journal Article Ji, X., Yuan, S., Li, J., Yin, P., Cao, H. & Xie, L. (2024). SGBA: semantic gaussian mixture model-based LiDAR bundle adjustment. IEEE Robotics and Automation Letters, 9(12), 10922-10929. https://dx.doi.org/10.1109/LRA.2024.3479699 2377-3766 https://hdl.handle.net/10356/182120 10.1109/LRA.2024.3479699 12 9 10922 10929 en IEEE Robotics and Automation Letters © 2024 IEEE. All rights reserved. This article may be downloaded for personal use only. Any other use requires prior permission of the copyright holder. The Version of Record is available online at http://doi.org/10.1109/LRA.2024.3479699. application/pdf |
| spellingShingle | Computer and Information Science Localization Mapping Bundle adjustment Ji, Xingyu Yuan, Shenghai Li, Jianping Yin, Pengyu Cao, Haozhi Xie, Lihua SGBA: semantic gaussian mixture model-based LiDAR bundle adjustment |
| title | SGBA: semantic gaussian mixture model-based LiDAR bundle adjustment |
| title_full | SGBA: semantic gaussian mixture model-based LiDAR bundle adjustment |
| title_fullStr | SGBA: semantic gaussian mixture model-based LiDAR bundle adjustment |
| title_full_unstemmed | SGBA: semantic gaussian mixture model-based LiDAR bundle adjustment |
| title_short | SGBA: semantic gaussian mixture model-based LiDAR bundle adjustment |
| title_sort | sgba semantic gaussian mixture model based lidar bundle adjustment |
| topic | Computer and Information Science Localization Mapping Bundle adjustment |
| url | https://hdl.handle.net/10356/182120 |
| work_keys_str_mv | AT jixingyu sgbasemanticgaussianmixturemodelbasedlidarbundleadjustment AT yuanshenghai sgbasemanticgaussianmixturemodelbasedlidarbundleadjustment AT lijianping sgbasemanticgaussianmixturemodelbasedlidarbundleadjustment AT yinpengyu sgbasemanticgaussianmixturemodelbasedlidarbundleadjustment AT caohaozhi sgbasemanticgaussianmixturemodelbasedlidarbundleadjustment AT xielihua sgbasemanticgaussianmixturemodelbasedlidarbundleadjustment |