Three-dimensional registration is an established yet challenging problem that is key in many different applications, such as mapping the environment for autonomous vehicles, or modeling people for avatar creation, among others. Registration refers ...

Domain adaptation (DA) is widely used in pattern recognition and computer vision. It learns a model from the source domain data and applies it in the target domain. Most of the DA techniques require access to source domain data. However, it could ...

Existing multi-person three-dimensional (3D) human pose estimation methods currently rely heavily on cameras, which are inevitably affected by smoke, haze, dust, poor lighting conditions, and privacy breaches in real-world scenarios. Due to the ...

Due to the annotation difficulty of point clouds, Unsupervised Domain Adaptation (UDA) is a promising direction to address unlabeled point cloud classification and segmentation. Recent works show that adding a self-supervised learning branch for ...

• We propose Graph Attentive Dual Ensemble (GRADE) for efficient semantic transfer in 3D point clouds.
• We propose a dual ensemble network for consistent generalization and accurate reconstruction.
• We propose a dynamic graph attentive ...

The circular hole structures on automotive engines possess stringent mechanical processing requirements, so it is of vital importance to perform quality inspections on all manufactured circular hole structures. The detection of circular holes on ...

Display Omitted

• A method to extract multi-scale and irregularly distributed circular holes in engine.
• By compartmentalization analysis, we enhance the observation of internal hole points.
• Curvature center contractility improves detection in small ...

In this paper, we adopt the Universal Manifold Embedding (UME) framework for the estimation of rigid transformations and extend it, so that it can accommodate scenarios involving partial overlap and differently sampled point clouds. UME is a ...${\mathrm{}}^{}\mathrm{}$$\mathrm{}$

Adversarial attacks on point clouds play a vital role in assessing and enhancing the adversarial robustness of 3D deep learning models. While employing a variety of geometric constraints, existing adversarial attack solutions often display ...

Autonomous driving demands high-quality LiDAR data, yet the cost of physical LiDAR sensors presents a significant scaling-up challenge. While recent efforts have explored deep generative models to address this issue, they often consume substantial ...

Currently, 3D object detection is a research hotspot in the field of computer vision. In this paper, we have observed that the commonly used set abstraction module retains excessive irrelevant background information during downsampling, impacting ...

• Proposes mixed sampling to enhance 3D object detection precision.
• Integrates semantic features for focused foreground point sampling.
• Introduces a module for robust feature extraction from high-quality 3D proposals.
• Achieves ...

In order to achieve truly immersive multimedia experiences, full freedom of movement has to be supported, and high-quality, interactive video delivery to the head-mounted device is vital. In wireless environments, this is very challenging due to ...

Due to the limited computational resources of the onboard computing devices of autonomous vehicles, the development of lightweight 3D object detectors is essential. Point-based detectors that progressively sample raw point clouds reduce numerous ...

• The negative effect of the points the road reflects on point sampling was studied.
• The proposed W-DFPS enhances the weight of foreground points in sampling results.
• A small number of sampling points can also cover most foreground ...

Efficiently detecting pedestrians from 3D point cloud data is a significantly challenging perception task in numerous robotic and autonomous driving applications, primarily because of the sparsity of point cloud data representing pedestrian ...

• We present an end-to-end trainable single-stage 3D pedestrian detection network.
• To retain spatial information lost during downsampling and suppress point cloud noise.
• The method ranks first on the JRDB 3D object detection ...

A template for representing basic building forms, termed 3BUGS, is proposed and demonstrated in a workflow fed with dense point clouds. The template describes building primitives using transformed rings for building geometry specification. Its ...

The rapid growth on the amount of generated 3D data, particularly in the form of Light Detection And Ranging (LiDAR) point clouds (PCs), poses very significant challenges in terms of data storage, transmission, and processing. Point cloud (PC) ...

In this case study, we investigate if it is possible to harness the capabilities of modern commodity hardware to perform ad-hoc queries on large raw geospatial data sets. Normally, this requires building an index structure, which is a time-...

Cross-modal setting employing 2D images and 3D point clouds in self-supervised representation learning is proven to be an effective way to enhance visual perception capabilities. However, different modalities have different data formats and ...

In this study, we introduce PlantSegNet, a novel neural network model for instance segmentation of nearby objects with similar geometric structures. Our work addresses the challenges of instance segmentation of plant point clouds, including the ...

Display Omitted

• A pipeline for generating procedural sorghum models, which we use to generate point clouds of sorghum fields.
• A large-scale, annotated, and synthetic dataset of sorghum plants.
• PlantSegNet model, a neural network structure for ...

The semantic segmentation of outdoor LiDAR point clouds is one of the gigantic fields in the large-scale driving scenario. However, the performances of the state-of-the-art methods are unsatisfactory caused by the intrinsic limitations of the ...

• V-SIAM enrichs voxel feature details and recalibrates the voxel features.
• MsVFE fuses multi-scale voxel information of the sparse points.
• Our methods achieve SOTA performances on Toronto3D and KITTI-360 datasets.

To support the roll-out of fibre broadband Internet in Germany, Deutsche Telekom has set itself the goal of connecting more than 2.5 million households per year to FTTH (Fibre to the Home). However, planning and approval processes have been very ...

• Software platform to speed up the fibre roll-out in Germany.
• Cloud-based, Big Data processing pipeline for 360°panorama imagery and 3D point clouds.
• Web-based 3D visualization used for interactive planning.
• More than 8 million ...

Self-supervised methods have been proven effective for learning deep representations of 3D point cloud data. Although recent methods in this domain often rely on random masking of inputs, the results of this approach can be improved. We introduce ...