IS-CAT: Intensity–Spatial Cross-Attention Transformer for LiDAR-Based Place Recognition
LiDAR place recognition is a crucial component of autonomous navigation, essential for loop closure in simultaneous localization and mapping (SLAM) systems. Notably, while camera-based methods struggle in fluctuating environments, such as weather or light, LiDAR demonstrates robustness against such...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-01-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/24/2/582 |
| _version_ | 1797339357986684928 |
|---|---|
| author | Hyeong-Jun Joo Jaeho Kim |
| author_facet | Hyeong-Jun Joo Jaeho Kim |
| author_sort | Hyeong-Jun Joo |
| collection | DOAJ |
| description | LiDAR place recognition is a crucial component of autonomous navigation, essential for loop closure in simultaneous localization and mapping (SLAM) systems. Notably, while camera-based methods struggle in fluctuating environments, such as weather or light, LiDAR demonstrates robustness against such challenges. This study introduces the intensity and spatial cross-attention transformer, which is a novel approach that utilizes LiDAR to generate global descriptors by fusing spatial and intensity data for enhanced place recognition. The proposed model leveraged a cross attention to a concatenation mechanism to process and integrate multi-layered LiDAR projections. Consequently, the previously unexplored synergy between spatial and intensity data was addressed. We demonstrated the performance of IS-CAT through extensive validation on the NCLT dataset. Additionally, we performed indoor evaluations on our Sejong indoor-5F dataset and demonstrated successful application to a 3D LiDAR SLAM system. Our findings highlight descriptors that demonstrate superior performance in various environments. This performance enhancement is evident in both indoor and outdoor settings, underscoring the practical effectiveness and advancements of our approach. |
| first_indexed | 2024-03-08T09:46:50Z |
| format | Article |
| id | doaj.art-a645d10a824f4aedb67878cac30fa7b4 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-08T09:46:50Z |
| publishDate | 2024-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-a645d10a824f4aedb67878cac30fa7b42024-01-29T14:16:44ZengMDPI AGSensors1424-82202024-01-0124258210.3390/s24020582IS-CAT: Intensity–Spatial Cross-Attention Transformer for LiDAR-Based Place RecognitionHyeong-Jun Joo0Jaeho Kim1Department of Information and Communications Engineering, Sejong University, Seoul 05006, Republic of KoreaDepartment of Electrical Engineering, Sejong University, Seoul 05006, Republic of KoreaLiDAR place recognition is a crucial component of autonomous navigation, essential for loop closure in simultaneous localization and mapping (SLAM) systems. Notably, while camera-based methods struggle in fluctuating environments, such as weather or light, LiDAR demonstrates robustness against such challenges. This study introduces the intensity and spatial cross-attention transformer, which is a novel approach that utilizes LiDAR to generate global descriptors by fusing spatial and intensity data for enhanced place recognition. The proposed model leveraged a cross attention to a concatenation mechanism to process and integrate multi-layered LiDAR projections. Consequently, the previously unexplored synergy between spatial and intensity data was addressed. We demonstrated the performance of IS-CAT through extensive validation on the NCLT dataset. Additionally, we performed indoor evaluations on our Sejong indoor-5F dataset and demonstrated successful application to a 3D LiDAR SLAM system. Our findings highlight descriptors that demonstrate superior performance in various environments. This performance enhancement is evident in both indoor and outdoor settings, underscoring the practical effectiveness and advancements of our approach.https://www.mdpi.com/1424-8220/24/2/582LiDAR place recognitionSLAMcross-attention transformer networkIS-CAT |
| spellingShingle | Hyeong-Jun Joo Jaeho Kim IS-CAT: Intensity–Spatial Cross-Attention Transformer for LiDAR-Based Place Recognition Sensors LiDAR place recognition SLAM cross-attention transformer network IS-CAT |
| title | IS-CAT: Intensity–Spatial Cross-Attention Transformer for LiDAR-Based Place Recognition |
| title_full | IS-CAT: Intensity–Spatial Cross-Attention Transformer for LiDAR-Based Place Recognition |
| title_fullStr | IS-CAT: Intensity–Spatial Cross-Attention Transformer for LiDAR-Based Place Recognition |
| title_full_unstemmed | IS-CAT: Intensity–Spatial Cross-Attention Transformer for LiDAR-Based Place Recognition |
| title_short | IS-CAT: Intensity–Spatial Cross-Attention Transformer for LiDAR-Based Place Recognition |
| title_sort | is cat intensity spatial cross attention transformer for lidar based place recognition |
| topic | LiDAR place recognition SLAM cross-attention transformer network IS-CAT |
| url | https://www.mdpi.com/1424-8220/24/2/582 |
| work_keys_str_mv | AT hyeongjunjoo iscatintensityspatialcrossattentiontransformerforlidarbasedplacerecognition AT jaehokim iscatintensityspatialcrossattentiontransformerforlidarbasedplacerecognition |