Rotation Invariant Spherical Harmonic Representation of 3D Shape Descriptors

By M. Kazhdan, Thomas A. Funkhouser, Szymon Rusinkiewicz. 2003

1,554 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

How researchers see this paper

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

In Computer Vision and Pattern Recognition

Citations in the corpus

Learning Descriptor Networks for 3D Shape Synthesis and Analysis (2018, neutral): “[23] 88.4% PANORAMA-NN [19] 91.1% ECC [22] 90.0% 3D ShapeNets [29] 83.5% DeepPano [21] 85.5% SPH [10] 79.8% VConv-DAE [20] 80.5% 3D-GAN [28] 91.0% 3D DescriptorNet (ours) 92.4 %”
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation (2016, background): “The test results for all six areas are aggregated for the PR curve generation.”
Opening the Black Box of Deep Neural Networks via Information (2017, methodology): “These orbits form a minimal sufﬁcient partition/statistics for spherically symmetric rules [Kazhdan et al. (2003)].”
Volumetric and Multi-view CNNs for Object Classification on 3D Data (2016, background): “The spherical harmonic descriptor (SPH) [15] and the Light Field descriptor (LFD) [4] are other popular descriptors.”
Multi-view Convolutional Neural Networks for 3D Shape Recognition (2015, background): “For our experiments, we use the same training and test split of ModelNet40 as in [37].2 Our shape descriptors are compared against the 3D ShapeNets by Wu et al. [37], the Spherical Harmonics descriptor (SPH) by Kazhdan et al. [16], the LightField descriptor (LFD) by Chen et al. [5], and Fisher vecto”
Multi-View Silhouette and Depth Decomposition for High Resolution 3D Object Representation (2018, background): “Other representations aim to use orientation [36], rotational invariance [15] or 3D-SURF features [16].”
Learning a Probabilistic Latent Space of Object Shapes via 3D Generative-Adversarial Modeling (2016, neutral): “…DeepPano [Shi et al., 2015] 77.6% 85.5% VoxNet [Maturana and Scherer, 2015] 83.0% 92.0% ORION [Sedaghat et al., 2016] - 93.8 % Unsupervised - SPH [Kazhdan et al., 2003] 68.2% 79.8% LFD [Chen et al., 2003] 75.5% 79.9% T-L Network [Girdhar et al., 2016] 74.4% - VConv-DAE [Sharma et al., 2016]…”
VConv-DAE: Deep Volumetric Shape Learning Without Object Labels (2016, methodology): “For comparison, we also report performance for Light Field descriptor (LFD [20], 4,700 dimensions) and Spherical Harmonic descriptor (SPH [21], 544 dimensions).”
FPNN: Field Probing Neural Networks for 3D Data (2016, background): “3D shapes can also be represented by their inherent statistical properties, such as distance distribution [22] and spherical harmonic decomposition [14].”
On Invariance and Selectivity in Representation Learning (2015, methodology): “For example in the analysis of shapes [25], and more generally in computational topology [14] or in the design of positive definite functions associated with reproducing kernel Hilbert spaces [18].”
FusionNet: 3D Object Classification Using Multiple Data Representations (2016, neutral): “Other shape descriptors include the Light Field Descriptor [3], Heat kernel signatures [2] [12] and SPH [11].”
PVNet: A Joint Convolutional Network of Point Cloud and Multi-View for 3D Shape Recognition (2018, background): “Classification Retrieval Pre train Fine tune #Number of Views (Overall Accuracy) (mAP) (1)SPH[16] 68.”
Scale-invariant heat kernel signatures for non-rigid shape recognition (2010, background): “Rotation and translation invariance can be achieved using volume and area descriptors [45], spherical harmonics [16], geometric moments et al. [38], and distribution of pair-wise Euclidean distances [29].”
Developing an engineering shape benchmark for CAD models (2006, background): “5D spherical harmonics [22], and 2D Shape Distributions [23,24] are some examples of view-based representations.”
GVCNN: Group-View Convolutional Neural Networks for 3D Shape Recognition (2018, methodology): “…by Qi et al. [18], which employs multi-resolution views, 3D ShapeNets by Wu et al. [26], Spherical Harmonics descriptor (SPH) by Kazhdan et al. [11], which is a typical model-based method, Lighting Field descriptor (LFD) by Chen et al. [4], which is a typical view-based method, PointNet by Qi…”
3D ShapeNets: A deep representation for volumetric shapes (2014, methodology): “For comparison, we choose Light Field descriptor [8] (LFD, 4,700 dimensions) and Spherical Harmonic descriptor [18] (SPH, 544 dimensions), which performed best among all descriptors [28].”
A survey of content based 3D shape retrieval methods (2004, methodology): “els that differ by a rotation to the interior part [43].”