Results 31 to 40 of about 5,826,707 (172)
TEASER: Fast and Certifiable Point Cloud Registration [PDF]
IEEE Transactions on robotics, 2020 We propose the first fast and certifiable algorithm for the registration of two sets of three-dimensional (3-D) points in the presence of large amounts of outlier correspondences.
Heng Yang, J. Shi, L. Carlone
semanticscholar +1 more source
CLIP2Point: Transfer CLIP to Point Cloud Classification with Image-Depth Pre-Training [PDF]
IEEE International Conference on Computer Vision, 2022 Pre-training across 3D vision and language remains under development because of limited training data. Recent works attempt to transfer vision-language (V-L) pre-training methods to 3D vision. However, the domain gap between 3D and images is unsolved, so
Tianyu Huang +6 more
semanticscholar +1 more source
NoiseTrans: Point Cloud Denoising with Transformers [PDF]
arXiv, 2023 Point clouds obtained from capture devices or 3D reconstruction techniques are often noisy and interfere with downstream tasks. The paper aims to recover the underlying surface of noisy point clouds. We design a novel model, NoiseTrans, which uses transformer encoder architecture for point cloud denoising.
arxiv
IDEA Journal, 2023 Laser scanning holds out the possibility of extreme certainty. Digital scanning has become deeply integrated in contemporary archaeological surveying, and in architectural heritage and preservation contexts digital scans are now common. Certainty in this
KeJia
semanticscholar +1 more source
SDFReg: Learning Signed Distance Functions for Point Cloud Registration [PDF]
arXiv, 2023 Learning-based point cloud registration methods can handle clean point clouds well, while it is still challenging to generalize to noisy, partial, and density-varying point clouds. To this end, we propose a novel point cloud registration framework for these imperfect point clouds.
arxiv
PointGuard: Provably Robust 3D Point Cloud Classification [PDF]
arXiv, 2021 3D point cloud classification has many safety-critical applications such as autonomous driving and robotic grasping. However, several studies showed that it is vulnerable to adversarial attacks. In particular, an attacker can make a classifier predict an incorrect label for a 3D point cloud via carefully modifying, adding, and/or deleting a small ...
arxiv
SpinNet: Learning a General Surface Descriptor for 3D Point Cloud Registration [PDF]
Computer Vision and Pattern Recognition, 2020 Extracting robust and general 3D local features is key to downstream tasks such as point cloud registration and reconstruction. Existing learning-based local descriptors are either sensitive to rotation transformations, or rely on classical handcrafted ...
Sheng Ao +4 more
semanticscholar +1 more source
APSIPA Transactions on Signal and Information Processing, 2020 This article presents an overview of the recent standardization activities for point cloud compression (PCC). A point cloud is a 3D data representation used in diverse applications associated with immersive media including virtual/augmented reality ...
D. Graziosi +5 more
semanticscholar +1 more source
Unsupervised Point Cloud Pre-training via Occlusion Completion [PDF]
IEEE International Conference on Computer Vision, 2020 We describe a simple pre-training approach for point clouds. It works in three steps: 1. Mask all points occluded in a camera view; 2. Learn an encoder-decoder model to reconstruct the occluded points; 3.
Hanchen Wang +4 more
semanticscholar +1 more source
MinkLoc3D: Point Cloud Based Large-Scale Place Recognition [PDF]
IEEE Workshop/Winter Conference on Applications of Computer Vision, 2020 The paper presents a learning-based method for computing a discriminative 3D point cloud descriptor for place recognition purposes. Existing methods, such as Point-NetVLAD, are based on unordered point cloud representation. They use PointNet as the first
J. Komorowski
semanticscholar +1 more source