Corpus ID: 21161163

Deep Spectral Descriptors: Learning the point-wise correspondence metric via Siamese deep neural networks

@article{Sun2017DeepSD,
  title={Deep Spectral Descriptors: Learning the point-wise correspondence metric via Siamese deep neural networks},
  author={Zhiyu Sun and Yusen He and Andrey Gritsenko and Amaury Lendasse and Stephen Seung-Yeob Baek},
  journal={ArXiv},
  year={2017},
  volume={abs/1710.06368}
}
A robust and informative local shape descriptor plays an important role in mesh registration. [...] Key Method We design and train a Siamese deep neural network to find such an embedding, where the embedded descriptors are promoted to rearrange based on the geometric similarity. We found our approach can significantly enhance the performance of the conventional spectral descriptors for the non-isometric registration tasks, and outperforms recent state-of-the-art method reported in literature.Expand
EdgeNet: Deep metric learning for 3D shapes
Abstract We introduce EdgeNet, a metric learning architecture for extracting semantic local shape features, directly applicable to a wide range of shape analysis applications such as point matching,Expand
ZerNet: Convolutional Neural Networks on Arbitrary Surfaces Via Zernike Local Tangent Space Estimation
TLDR
It is proved that the convolution of two functions can be represented as a simple dot product between Zernike coefficients and the rotation of a convolution kernel is essentially a set of 2 × 2 rotation matrices applied to the coefficients. Expand
Deep Geometric Functional Maps: Robust Feature Learning for Shape Correspondence
TLDR
This work presents a novel learning-based approach for computing correspondences between non-rigid 3D shapes that can learn from less training data than existing supervised approaches and generalizes significantly better than current descriptor-based learning methods. Expand
Free-Form Feature Classification for Finite Element Meshing based on Shape Descriptors and Machine Learning
Finite element mesh generation (FE meshing) from three-dimensional (3D) computer-aided design (CAD) models is generally the most critical process in the finite element analysis pipeline. In the FEExpand
Body shape matters: Evidence from machine learning on body shape-income relationship
TLDR
Estimation results reveal a statistically significant relationship between physical appearance and family income and that these associations differ across genders, which supports the hypothesis on the physical attractiveness premium in labor market outcomes and its heterogeneity across genders. Expand
Bringing Interpretability and Visualization with Artificial Neural Networks.
TLDR
Extensions of ELMs for non-typical for Artificial Neural Networks (ANNs) problems are presented, and an alternative way of interpreting probabilistic outputs for multi-class classification problems is proposed. Expand
Modeling and Optimizing Building HVAC Energy Systems Using Deep Neural Networks
  • Jiahao Deng, Haoran Wang
  • Environmental Science
  • 2018 International Conference on Smart Grid and Clean Energy Technologies (ICSGCE)
  • 2018
The heating, ventilating and air conditioning (HVAC) systems consumes more than half of the building energy consumption. An efficient HVAC system ensures the comfortable living and workingExpand
Risk Assessment of Groundwater Depletion Induced Land Subsidence: A Case Study in Taiyuan Basin, China
Groundwater depletion induced land subsidence affects the safety of local communities. In this research, a data-driven approach is applied to predict and assess the risk level of land subsidence inExpand
Predictive Modeling of Mining Induced Ground Subsidence with Survival Analysis and Online Sequential Extreme Learning Machine
Mining induced land subsidence is one of the most hazardous geological phenomenon. Predictive modeling of the ground subsidence has attracted increased interest and is crucial to the hazardExpand

References

SHOWING 1-10 OF 49 REFERENCES
Discriminative Learning of Deep Convolutional Feature Point Descriptors
TLDR
This paper uses Convolutional Neural Networks to learn discriminant patch representations and in particular train a Siamese network with pairs of (non-)corresponding patches to develop 128-D descriptors whose euclidean distances reflect patch similarity and can be used as a drop-in replacement for any task involving SIFT. Expand
Learning shape correspondence with anisotropic convolutional neural networks
TLDR
An intrinsic convolutional neural network architecture based on anisotropic diffusion kernels is introduced, which is term Anisotropic Convolutional Neural Network (ACNN), and is used to effectively learn intrinsic dense correspondences between deformable shapes in very challenging settings. Expand
Spatially aggregating spectral descriptors for nonrigid 3D shape retrieval: a comparative survey
TLDR
This paper presents a comprehensive review and analysis of recent spectral shape descriptors for nonrigid 3D shape retrieval, and proposes to adopt the isocontours of the second eigenfunction of the LB operator to perform surface partition, which can significantly ameliorate the retrieval performance of the time-scaled local descriptors. Expand
Optimal Intrinsic Descriptors for Non-Rigid Shape Analysis
TLDR
Novel point descriptors for 3D shapes with the potential to match two shapes representing the same object undergoing natural deformations are proposed, and they are guaranteed to be at least as precise as any Heat Kernel Signature or Wave Kernel Signature of any parameterisation. Expand
Learning Local Image Descriptors with Deep Siamese and Triplet Convolutional Networks by Minimizing Global Loss Functions
TLDR
A combination of the triplet and global losses produces the best embedding in the field, using this triplet network, and it is demonstrated that the use of the central-surround siamese network trained with the global loss producing the best result of the field on the UBC dataset. Expand
Geodesic Convolutional Neural Networks on Riemannian Manifolds
TLDR
Geodesic Convolutional Neural Networks (GCNN), a generalization of the convolutional neural networks (CNN) paradigm to non-Euclidean manifolds is introduced, allowing to achieve state-of-the-art performance in problems such as shape description, retrieval, and correspondence. Expand
Supervised Descriptor Learning for Non-Rigid Shape Matching
TLDR
A novel method for computing correspondences between pairs of non-rigid shapes by considering the problem of learning a correspondence model given a collection of reference pairs with known mappings between them and using the recently proposed functional maps framework. Expand
Scale-invariant heat kernel signatures for non-rigid shape recognition
  • M. Bronstein, I. Kokkinos
  • Mathematics, Computer Science
  • 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition
  • 2010
TLDR
A scale-invariant version of the heat kernel descriptor that can be used in the bag-of-features framework for shape retrieval in the presence of transformations such as isometric deformations, missing data, topological noise, and global and local scaling. Expand
PN-Net: Conjoined Triple Deep Network for Learning Local Image Descriptors
TLDR
This paper proposes a CNN based descriptor with improved matching performance, significantly reduced training and execution time, as well as low dimensionality, and introduces a new loss function that exploits the relations within the triplets. Expand
Dense Human Body Correspondences Using Convolutional Networks
TLDR
This work uses a deep convolutional neural network to train a feature descriptor on depth map pixels, but crucially, rather than training the network to solve the shape correspondence problem directly, it trains it to solve a body region classification problem, modified to increase the smoothness of the learned descriptors near region boundaries. Expand
...
1
2
3
4
5
...