Deep regression for LiDAR-based localization in dense urban areas

Yu, Shangshu; Wang, Cheng; Yu, Zenglei; Li, Xin; Cheng, Ming; Zang, Yu

doi:10.1016/j.isprsjprs.2020.12.013

Cited by 16 publications

(4 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Following (Yu et al, 2021), we use the relocalization recall as the evaluation metric. The relocalization recall denotes the percentage of query samples with regression pose error under certain thresholds.…”

Section: Experiments and Evaluationmentioning

confidence: 99%

Relocalization based on millimeter wave radar point cloud for visually degraded environments

Cheng

Pang

Jiang³

et al. 2023

Journal of Field Robotics

View full text Add to dashboard Cite

Simultaneously localization and mapping (SLAM) has been widely used in autonomous mobile systems to fulfill autonomous navigation. Relocalization plays an important role in SLAM for closing the loop and eliminating the drift of pose estimation. Traditional methods mostly rely on LiDAR or camera sensors, which may degrade or even fail in rainy or dusty situations or with large illumination changes. In this article, we explore the use of low‐cost commercial millimeter wave (mmWave) radars and propose a noval mmWave radar point cloud‐based relocalization method. Our method first pre‐processes the radar point cloud and, based on that, achieves fast 3‐DOF pose estimation for the robot. We build a prototype and thoroughly evaluate our method using data sets collected by our platform in four complex environments, including street, park, road, and water surface scenarios. The experimental results show that our method consistently outperforms other baseline methods including the vision‐based counterparts, especially in the visual degraded scenes.

show abstract

Section: Experiments and Evaluationmentioning

confidence: 99%

Relocalization based on millimeter wave radar point cloud for visually degraded environments

Cheng

Pang

Jiang³

et al. 2023

Journal of Field Robotics

View full text Add to dashboard Cite

show abstract

“…Therefore, SLAM-based techniques are more suitable for AV applications that involve navigating in an unknown environment. 11 By contrast, HD mapping methods provide a detailed map of the surrounding environment to assist the AV in its perception. With this approach, it is much simpler to identify objects, since their information is already available in the database.…”

Section: Introductionmentioning

confidence: 99%

High-definition 3D map for autonomous vehicles in Vietnam: A comparison between graph simultaneous localization and mapping and the normal distributions transform algorithm

Nam

Son

2023

International Journal of Advanced Robotic Systems

View full text Add to dashboard Cite

Autonomous vehicles have a fundamental part to play in future transportation. In order to work, such vehicles need a way of determining their exact location. To date, two techniques have proved applicable: geolocation using GPS and 3D mapping. In the latter case, the creation of the 3D map is an essential step in preparing the software. This manuscript presents a novel method for creating a high-definition 3D map using a Velodyne sensor. The primary objective is to achieve deeper understanding of the map creation algorithm and to lay the groundwork for future autonomous vehicle architecture. In this manuscript, two high-definition mapping algorithms are applied: the graph simultaneous localization and mapping algorithm and the normal distributions transform algorithm. The results of both displayed different strengths and weaknesses upon creating a high-definition 3D map, which may help other researchers to select the best algorithm in future.

show abstract

“…One specific aspect of localization is place recognition, which involves searching a database of geo-tagged scene data to find the descriptor most similar to the query scene. This becomes particularly important when localizing in environments where reliable GPS signals are unavailable [1] or when using SLAM systems that require loop closure [2][3][4]. While vision-based methods can be sensitive to factors like illumination, camera field of view, and viewing orientation [5], lidar-based solutions offer greater robustness to different lighting conditions and seasonal changes.…”

Section: Introductionmentioning

confidence: 99%

LWR-Net: Robust and Lightweight Place Recognition Network for Noisy and Low-Density Point Clouds

Zhang,

Chen,

Shu

et al. 2023

Sensors

View full text Add to dashboard Cite

Point cloud-based retrieval for place recognition is essential in robotic applications like autonomous driving or simultaneous localization and mapping. However, this remains challenging in complex real-world scenes. Existing methods are sensitive to noisy, low-density point clouds and require extensive storage and computation, posing limitations for hardware-limited scenarios. To overcome these challenges, we propose LWR-Net, a lightweight place recognition network for efficient and robust point cloud retrieval in noisy, low-density conditions. Our approach incorporates a fast dilated sampling and grouping module with a residual MLP structure to learn geometric features from local neighborhoods. We also introduce a lightweight attentional weighting module to enhance global feature representation. By utilizing the Generalized Mean pooling structure, we aggregated the global descriptor for point cloud retrieval. We validated LWR-Net’s efficiency and robustness on the Oxford robotcar dataset and three in-house datasets. The results demonstrate that our method efficiently and accurately retrieves matching scenes while being more robust to variations in point density and noise intensity. LWR-Net achieves state-of-the-art accuracy and robustness with a lightweight model size of 0.4M parameters. These efficiency, robustness, and lightweight advantages make our network highly suitable for robotic applications relying on point cloud-based place recognition.

show abstract

Deep regression for LiDAR-based localization in dense urban areas

Cited by 16 publications

References 16 publications

Relocalization based on millimeter wave radar point cloud for visually degraded environments

Relocalization based on millimeter wave radar point cloud for visually degraded environments

High-definition 3D map for autonomous vehicles in Vietnam: A comparison between graph simultaneous localization and mapping and the normal distributions transform algorithm

LWR-Net: Robust and Lightweight Place Recognition Network for Noisy and Low-Density Point Clouds

Contact Info

Product

Resources

About