Deep Voxelized Feature Maps for Self-Localization in Autonomous Driving
Lane-level self-localization is essential for autonomous driving. Point cloud maps are typically used for self-localization but are known to be redundant. Deep features produced by neural networks can be used as a map, but their simple utilization could lead to corruption in large environments. This...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-06-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/23/12/5373 |
| _version_ | 1797592709597233152 |
|---|---|
| author | Yuki Endo Shunsuke Kamijo |
| author_facet | Yuki Endo Shunsuke Kamijo |
| author_sort | Yuki Endo |
| collection | DOAJ |
| description | Lane-level self-localization is essential for autonomous driving. Point cloud maps are typically used for self-localization but are known to be redundant. Deep features produced by neural networks can be used as a map, but their simple utilization could lead to corruption in large environments. This paper proposes a practical map format using deep features. We propose voxelized deep feature maps for self-localization, consisting of deep features defined in small regions. The self-localization algorithm proposed in this paper considers per-voxel residual and reassignment of scan points in each optimization iteration, which could result in accurate results. Our experiments compared point cloud maps, feature maps, and the proposed map from the self-localization accuracy and efficiency perspective. As a result, more accurate and lane-level self-localization was achieved with the proposed voxelized deep feature map, even with a smaller storage requirement compared with the other map formats. |
| first_indexed | 2024-03-11T01:57:59Z |
| format | Article |
| id | doaj.art-633dce77522a4c198f32c0a489e8f70f |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-11T01:57:59Z |
| publishDate | 2023-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-633dce77522a4c198f32c0a489e8f70f2023-11-18T12:30:05ZengMDPI AGSensors1424-82202023-06-012312537310.3390/s23125373Deep Voxelized Feature Maps for Self-Localization in Autonomous DrivingYuki Endo0Shunsuke Kamijo1Department of Information & Communication Engineering, Graduate School of Information Science and Technology, The University of Tokyo, Tokyo 153-8505, JapanThe Institute of Industrial Science (IIS), The University of Tokyo, Tokyo 153-8505, JapanLane-level self-localization is essential for autonomous driving. Point cloud maps are typically used for self-localization but are known to be redundant. Deep features produced by neural networks can be used as a map, but their simple utilization could lead to corruption in large environments. This paper proposes a practical map format using deep features. We propose voxelized deep feature maps for self-localization, consisting of deep features defined in small regions. The self-localization algorithm proposed in this paper considers per-voxel residual and reassignment of scan points in each optimization iteration, which could result in accurate results. Our experiments compared point cloud maps, feature maps, and the proposed map from the self-localization accuracy and efficiency perspective. As a result, more accurate and lane-level self-localization was achieved with the proposed voxelized deep feature map, even with a smaller storage requirement compared with the other map formats.https://www.mdpi.com/1424-8220/23/12/5373autonomous drivingself-localizationdeep learning |
| spellingShingle | Yuki Endo Shunsuke Kamijo Deep Voxelized Feature Maps for Self-Localization in Autonomous Driving Sensors autonomous driving self-localization deep learning |
| title | Deep Voxelized Feature Maps for Self-Localization in Autonomous Driving |
| title_full | Deep Voxelized Feature Maps for Self-Localization in Autonomous Driving |
| title_fullStr | Deep Voxelized Feature Maps for Self-Localization in Autonomous Driving |
| title_full_unstemmed | Deep Voxelized Feature Maps for Self-Localization in Autonomous Driving |
| title_short | Deep Voxelized Feature Maps for Self-Localization in Autonomous Driving |
| title_sort | deep voxelized feature maps for self localization in autonomous driving |
| topic | autonomous driving self-localization deep learning |
| url | https://www.mdpi.com/1424-8220/23/12/5373 |
| work_keys_str_mv | AT yukiendo deepvoxelizedfeaturemapsforselflocalizationinautonomousdriving AT shunsukekamijo deepvoxelizedfeaturemapsforselflocalizationinautonomousdriving |