Semantic Graph Based Place Recognition for 3D Point Clouds

Kong, Xin; Yang, Xin; Zhai, Guangyao; Zhao, Xiangrui; Zeng, Xianfang; Wang, Mengmeng; Liu, Yong; Li, Wanlong; Wen, Feng

doi:10.1109/iros45743.2020.9341060

Cited by 106 publications

(64 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Semantic information was first exploited in place recognition tasks to improve the ability of scene representation. SGPR [17] and SSC [18] both used semantic information of the points to describe the scene. SGPR segmented the point cloud in different semantic classes and used the Euclidean clustering in each class to get more instances.…”

Section: Semantic-based Methodsmentioning

confidence: 99%

“…Since centroid coordinates alone are not representive, more descriptions need to be extracted for the segments. SGPR [17] used the pre-trained semantic segmentation network to classify the segments. However, they are unable to distinguish the difference between segments of the same class.…”

Section: B Segments Feature Extractionmentioning

confidence: 99%

“…the arrangement of certain segments and their spatial relations, hence it is difficult to distinguish scenarios with similar features but different structures. In order to get distinct features, high-level representations such as geometrical segments or semantics were proposed [14], [15], [16], [17], [18]. SegMatch [14] first used Euclidean clustering to segment a point cloud and search candidates by matching K-nearest neighbour (KNN) segments.…”

Section: Introductionmentioning

confidence: 99%

“…SGPR [17] introduced the semantic information to achieve a stable representation. SSC [18] further employed iterative closet point (ICP) and semantics to align the point cloud pairs to get more accurate matching results.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

DSC: Deep Scan Context Descriptor for Large-Scale Place Recognition

Cui¹,

Huang²,

Cai³

et al. 2021

Preprint

View full text Add to dashboard Cite

LiDAR-based place recognition is an essential and challenging task both in loop closure detection and global relocalization. We propose Deep Scan Context (DSC), a general and discriminative global descriptor that captures the relationship among segments of a point cloud. Unlike previous methods that utilize either semantics or a sequence of adjacent point clouds for better place recognition, we only use raw point clouds to get competitive results. Concretely, we first segment the point cloud egocentrically to acquire centroids and eigenvalues of the segments. Then, we introduce a graph neural network to aggregate these features into an embedding representation. Extensive experiments conducted on the KITTI dataset show that DSC is robust to scene variants and outperforms existing methods.

show abstract

Section: Semantic-based Methodsmentioning

confidence: 99%

Section: B Segments Feature Extractionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

DSC: Deep Scan Context Descriptor for Large-Scale Place Recognition

Cui¹,

Huang²,

Cai³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Most recently, Zaganidis et al (2019) proposed a normal distributions transform histogrambased loop closure detection method, which is assisted by semantic information. Kong et al (2020) also use semantic graphs for place recognition for 3D point clouds. Their network is capable of capturing topological and semantic information from the point cloud and also achieves rotational invariance.…”

Section: Loop Closing For Slammentioning

confidence: 99%

OverlapNet: a siamese network for computing LiDAR scan similarity with applications to loop closing and localization

et al. 2021

View full text Add to dashboard Cite

Localization and mapping are key capabilities of autonomous systems. In this paper, we propose a modified Siamese network to estimate the similarity between pairs of LiDAR scans recorded by autonomous cars. This can be used to address both, loop closing for SLAM and global localization. Our approach utilizes a deep neural network exploiting different cues generated from LiDAR data. It estimates the similarity between pairs of scans using the concept of image overlap generalized to range images and furthermore provides a relative yaw angle estimate. Based on such predictions, our method is able to detect loop closures in a SLAM system or to globally localize in a given map. For loop closure detection, we use the overlap prediction as the similarity measurement to find loop closure candidates and integrate the candidate selection into an existing SLAM system to improve the mapping performance. For global localization, we propose a novel observation model using the predictions provided by OverlapNet and integrate it into a Monte-Carlo localization framework. We evaluate our approach on multiple datasets collected using different LiDAR scanners in various environments. The experimental results show that our method can effectively detect loop closures surpassing the detection performance of state-of-the-art methods and that it generalizes well to different environments. Furthermore, our method reliably localizes a vehicle in typical urban environments globally using LiDAR data collected in different seasons.

show abstract

AttDLNet: Attention-Based Deep Network for 3D LiDAR Place Recognition

Barros

Garrote

Pereira

et al. 2022

Lecture Notes in Networks and Systems

View full text Add to dashboard Cite

Robust and reliable place recognition and loop closure detection in agricultural environments is still an open problem. In particular, orchards are a difficult case study due to structural similarity across the entire field. In this work, we address the place recognition problem in orchards resorting to 3D LiDAR data, which is considered a key modality for robustness. Hence, we propose ORCHNet, a deep-learning-based approach that maps 3D-LiDAR scans to global descriptors. Specifically, this work proposes a new global feature aggregation approach, which fuses multiple aggregation methods into a robust global descriptor. ORCHNet is evaluated on real-world data collected in orchards, comprising data from the summer and autumn seasons. To assess the robustness, We compare ORCHNet with state-ofthe-art aggregation approaches on data from the same season and across seasons. Moreover, we additionally evaluate the proposed approach as part of a localization framework, where ORCHNet is used as a loop closure detector. The empirical results indicate that, on the place recognition task, ORCHNet outperforms the remaining approaches, and is also more robust across seasons. As for the localization, the edge cases where the path goes through the trees are solved when integrating ORCHNet as a loop detector, showing the potential applicability of the proposed approach in this task. The code and dataset will be publicly available at:https://github.com/Cybonic/ORCHNet.git

show abstract

Semantic Graph Based Place Recognition for 3D Point Clouds

Cited by 106 publications

References 30 publications

DSC: Deep Scan Context Descriptor for Large-Scale Place Recognition

DSC: Deep Scan Context Descriptor for Large-Scale Place Recognition

OverlapNet: a siamese network for computing LiDAR scan similarity with applications to loop closing and localization

AttDLNet: Attention-Based Deep Network for 3D LiDAR Place Recognition

Contact Info

Product

Resources

About