Unique Signatures of Histograms for Local Surface Description

By Federico Tombari, Samuele Salti, Luigi Di Stefano. 2010

1,636 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Computational Mechanics, Computer Vision and Pattern Recognition, Geology, 3D Vision & Pose

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Aerospace Engineering

In Computational Mechanics

In Computer Vision and Pattern Recognition

In Geology

In 3D Vision & Pose

Citations in the corpus

Geometric Deep Learning on Graphs and Manifolds Using Mixture Model CNNs (2016, methodology): “For all the shapes, point-wise 544-dimensional SHOT descriptors (local histogram of normal vectors) [39] were used as input data.”
Geometric Deep Learning: Going beyond Euclidean data (2016, methodology): “SHOT descriptors capturing the local normal vector orientations [110] were used in this example as input features.”
SplineCNN: Fast Geometric Deep Learning with Continuous B-Spline Kernels (2017, methodology): “[18, 2, 17, 16], we go without handcrafted feature descriptors as inputs, like the local histogram of normal vectors known as SHOT descriptors [24], and force the network to learn from the geometry (i.”
Deep Functional Maps: Structured Prediction for Dense Shape Correspondence (2017, methodology): “For this test we use FAUST templates [7] as our data and SHOT [38] as an input feature.”
Deformable Shape Completion with Graph Convolutional Autoencoders (2017, background): “In an attempt to ﬁnd dense correspondences between the point cloud extracted from the Kinect sensor and a model of a human, we have noticed that off-the-shelf descriptors such as SHOT [53], HKS [51] and WKS [3], did not transfer well between the partial point cloud and the full mesh domains.”
DeepIM: Deep Iterative Matching for 6D Pose Estimation (2018, background): “2016), spin-images (Johnson and Hebert 1999), and point-pair histograms (Rusu et al. 2009; Tombari et al. 2010).”
Robotic Tactile Perception of Object Properties: A Review (2017, neutral): “As both visual images and tactile readings are present in numerical matrices, many other vision descriptors [104], [105] also have the potential to be applied in tactile sensing, e.g., Shape Context [106], SURF [107] and 3D descriptor SHOT [108].”
SHOT: Unique signatures of histograms for surface and texture description (2014, methodology): “In each test, we use the three datasets proposed in [13], which can be downloaded from the SHOT project page.”
SSD-6D: Making RGB-Based 3D Detection and 6D Pose Estimation Great Again (2017, methodology): “Rotation-invariant methods are based on local keypoints in either 2D [32] or 3D [7, 3, 30] by explicitly computing or voting for an orientation or a local reference frame, but they fail for objects of poor geometry or texture.”
DiffusionNet: Discretization Agnostic Learning on Surfaces (2020, background): “Higherorder descriptors such as SHOT [Tombari et al. 2010] seem unnecessary and may be unstable under remeshing [Donati et al.”
CoFiNet: Reliable Coarse-to-fine Correspondences for Robust Point Cloud Registration (2021, methodology): “From early-stage hand-crafted methods [1, 2, 3, 4, 5] to recently emerged deep learning-based approaches [6, 7, 8, 9, 10, 11, 12], many works contributed to improving the reliability of correspondences.”
Performance Evaluation of 3D Keypoint Detectors (2012, methodology): “(2010), acquired with a laser scanner; the dataset used for the experimental validation in Tombari et al. (2010), obtained by means of the SpaceTime Stereo acquisition technique; a novel dataset, acquired in our laboratory by a Microsoft Kinect device.”
A Global Hypotheses Verification Method for 3D Object Recognition (2012, methodology): “Then, the SHOT local descriptor [2] is computed at each keypoint over a support size specified by radius σd.”
3DMatch: Learning Local Geometric Descriptors from RGB-D Reconstructions (2016, methodology): “For Johnson et al. (Spin-Images) [19] and Rusu et al. (Fast Point Feature Histograms) [28], we use the implementation provided in the Point Cloud Library (PCL).”
Unique shape context for 3d data description (2010, methodology): “The definition of a unique, unambiguous local RF has been lacking in literature until recently [18].”
T-LESS: An RGB-D Dataset for 6D Pose Estimation of Texture-Less Objects (2017, background): “Instead, recent approaches that can deal with texture-less objects have focused on local 3D feature description [33, 51, 19], and semi-global or 2017 IEEE Winter Conference on Applications of Computer Vision”
3DmFV: Three-Dimensional Point Cloud Classification in Real-Time Using Convolutional Neural Networks (2018, background): “They can be divided into two main groups: local descriptors [8], [9], [10], [11] and global descriptors [12], [13], [14].”
PointNetVLAD: Deep Point Cloud Based Retrieval for Large-Scale Place Recognition (2018, background): “However, both [42, 45] rely on stable surfaces for descriptor calculation and are more suitable for dense rigid objects from 3D range images but not for outdoor LiDAR scans.”
Pointwise Rotation-Invariant Network with Adaptive Sampling and 3D Spherical Voxel Convolution (2018, methodology): “The development of rotation-invariant features from geometries could be retrospected to manual designed features, including Structural Indexing (Stein and Medioni 1992), Signature of Histogram Orientations (SHOT) (Tombari, Salti, and Di Stefano 2010) and CGF (Khoury, Zhou, and Koltun 2017).”