Faster 3D Object Detection in RGB-D Image Using 3D Selective Search and Object Pruning

Jiang, Liu; Chen, Hongliang; Li, Jianxun

doi:10.1109/ccdc.2018.8407973

Cited by 6 publications

(9 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The stage 1 of our previous work PointTrans [32] is introduced to generate instance-wise masks. Point-Voxel Convolution [26] is adopted to extract features from raw points.…”

Section: Instance Descriptor Generationmentioning

confidence: 99%

“…Therefore, We consider applying a density-based cluster method, DBSCAN [19], to address this problem. As it can be seen from the Figure 2, individual instances can be segmented by DBSCAN directly [32].…”

Section: Instance Descriptor Generationmentioning

confidence: 99%

“…We ingeniously take advantages of the separability of 3D point clouds and combination 3D semantic segmentation method with a density-based cluster method to directly generate masks for every object instance. However, this also means that the detection result of our algorithm will, to some extent, rely on the segmentation result of the semantic segmentation stage [32]. (1) Semantic segmentation and instance descriptor generation.…”

Section: Instance Descriptor Generationmentioning

confidence: 99%

“…Next, the static object points are filtered by selecting the specified semantic labels. The DBSCAN algorithm does not require the exact number of object instances [32]. Different parameters of DBSCAN cluster are set for different classes of object instances.…”

Section: Dataset and Trainingmentioning

confidence: 99%

See 3 more Smart Citations

PCRMLP : A Two-Stage Network for Point Cloud Registration in Urban Scenes

Liu¹,

Xu²,

Zhang³

et al. 2023

Preprint

View full text Add to dashboard Cite

Urban scene point cloud pose significant challenges for registration due to its large data volume, similar scenarios and dynamic objects. In this paper, we propose PCRMLP, a model for urban scene point cloud registration that achieves comparable registration performance to prior learning-based methods. Compared to previous works which focus on extracting features and estimating correspondence, the model estimates the transformation implicitly from concrete instances. An instance-level urban scene representation method is introduced to extract instance descriptors via semantic segmentation and DBSCAN, which enable the model to obtain robust instance features, filter dynamic objects and estimate transformation in a more logical manner. Then a lightweight network consisting of MLPs is employed to obtain transformation in an encoder-decoder manner. We validate the approach on KITTI dataset. Experimental results demonstrate that PCRMLP can obtain a satisfactory coarse transformation from instance descriptors just in 0.0028s. With a subsequent ICP refinement module, the proposed method achieves higher registration accuracy and computational efficiency than prior learning-based works.

show abstract

“…The stage 1 of our previous work PointTrans [32] is introduced to generate instance-wise masks. Point-Voxel Convolution [26] is adopted to extract features from raw points.…”

Section: Instance Descriptor Generationmentioning

confidence: 99%

Section: Instance Descriptor Generationmentioning

confidence: 99%

Section: Instance Descriptor Generationmentioning

confidence: 99%

Section: Dataset and Trainingmentioning

confidence: 99%

See 2 more Smart Citations

PCRMLP : A Two-Stage Network for Point Cloud Registration in Urban Scenes

Liu¹,

Xu²,

Zhang³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Alike, Liu et al (2018a) also use SVMs learned for each object class based on the feature selection proposed by Ren and Sudderth (2016). Through a pruning of candidates, by comparing the cuboid size of the bounding boxes with the distribution of the physical size of the objects, they further reduce inference time of detection.…”

Section: Sliding Window Approachesmentioning

confidence: 99%

Survey and Systematization of 3D Object Detection Models and Methods

Moritz¹,

Friederich²,

Egger³

et al. 2022

Preprint

View full text Add to dashboard Cite

This paper offers a comprehensive survey of recent developments in 3D object detection covering the full pipeline from input data, over data representation and feature extraction to the actual detection modules. We include basic concepts, focus our survey on a broad spectrum of different approaches arising in the last ten years and propose a systematization which offers a practical framework to compare those approaches on the methods level.

show abstract