In this work, we propose a novel approach to learn different non-rigid transformations of the input point cloud for different local neighborhoods at each layer. We propose both linear (affine) and non-linear (projective and deformable) spatial transformer for 3D points. With spatial transformers ...
Spatial-Temporal Transformer for 3D Point Cloud Sequences Yimin Wei1,2, Hao Liu1,2, Tingting Xie3, Qiuhong Ke4, Yulan Guo1,2,∗ 1Sun Yat-sen University, 2Shenzhen Campus of Sun Yat-sen University, 3Queen Mary University of London, 4The University of Melbourne weiym...
Leng, Z, Sun P, He T, Anguelov D, Tan M (2024) Pvtransformer: Point-to-voxel transformer for scalable 3d object detection. arXiv preprint arXiv:2405.02811 Wu X, Jiang L, Wang P-S, Liu Z, Liu X, Qiao Y, Ouyang W, He T, Zhao H (2024) Point transformer v3: Simpler faster str...
SMIFormer: Learning Spatial Feature Representation for 3D Object Detection from 4D Imaging Radar via Multi-View Interactive Transformersdoi:10.3390/s23239429OBJECT recognition (Computer vision)TRANSFORMER modelsPOINT cloudMILLIMETER wavesAUTONOMOUS vehicles...
it is imperative to reconstruct incomplete 3D point clouds of plants collected in complex planting scenarios. However, existing methods for plant point cloud completion mainly focus on the extraction of global features and exhibit limitations in effectively aggregating local and global features well, whic...
Semantic segmentation of point cloud data of architectural cultural heritage is of significant importance for HBIM modeling, disease extraction and analysis, and heritage restoration research fields. In the semantic segmentation task of architectural poi
transformer model for grounding 3D objects and their spatial relations. To this end, we design a spatial self-attention layer that accounts for relative distances and orientations between objects in input 3D point clouds. Training such a layer with visual and language inputs enables to disambiguate...
Please use the following command for installation: #1. It is recommended to create a new environmentconda create -n geotransformer python==3.8 conda activate geotransformer#2. Install vision3d following https://github.com/qinzheng93/vision3d ...
RV is the native represen- tation of LiDAR point clouds, therefore, it can produce compact and dense features. However, projection would inevitably impair the integrity of spatial information con- veyed in the 3D space no matter which of BEV or RV is chosen...
TSPconv-Net: Transformer and Sparse Convolution for 3D Instance Segmentation in Point Clouds Current deep learning approaches for indoor 3D instance segmentation often rely on multilayer perceptrons (MLPs) for feature extraction. However, MLPs struggle to effectively capture the complex spatial ...