3D-CODED: 3D Correspondences by Deep Deformation (2018-06-13T00:00:00.000000Z)

TL;DR

This work presents a new deep learning approach for matching deformable shapes by introducing Shape Deformation Networks which jointly encode 3D shapes and correspondences, and shows that this method is robust to many types of perturbations, and generalizes to non-human shapes.

Abstract

We present a new deep learning approach for matching deformable shapes by introducing Shape Deformation Networks which jointly encode 3D shapes and correspondences. This is achieved by factoring the surface representation into (i) a template, that parameterizes the surface, and (ii) a learnt global feature vector that parameterizes the transformation of the template into the input surface. By predicting this feature for a new shape, we implicitly predict correspondences between this shape and the template. We show that these correspondences can be improved by an additional step which improves the shape feature by minimizing the Chamfer distance between the input and transformed template. We demonstrate that our simple approach improves on state-of-the-art results on the difficult FAUST-inter challenge, with an average correspondence error of 2.88 cm. We show, on the TOSCA dataset, that our method is robust to many types of perturbations, and generalizes to non-human shapes. This robustness allows it to perform well on real unclean, meshes from the SCAPE dataset.

Authors

Bryan C. Russell

13 papers

Matthew Fisher

6 papers

Mathieu Aubry

9 papers

TL;DR

Abstract

Authors

References50 items

Lions and Tigers and Bears: Capturing Non-rigid, 3D, Articulated Shape from Images

Learning Category-Specific Mesh Reconstruction from Image Collections

A Papier-Mache Approach to Learning 3D Surface Generation

FoldingNet: Point Cloud Auto-Encoder via Deep Grid Deformation

Position-based tensegrity design

GWCNN: A Metric Alignment Layer for Deep Shape Analysis

Convolutional neural networks on surfaces via seamless toric covers

PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space

Deep Functional Maps: Structured Prediction for Dense Shape Correspondence

SurfNet: Generating 3D Shape Surfaces Using Deep Residual Networks

Learning from Synthetic Humans

PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation

A Point Set Generation Network for 3D Object Reconstruction from a Single Image

Geometric Deep Learning on Graphs and Manifolds Using Mixture Model CNNs

3D Menagerie: Modeling the 3D Shape and Pose of Animals

Deep Learning 3D Shape Surfaces Using Geometry Images

Entropic metric alignment for correspondence problems

Learning shape correspondence with anisotropic convolutional neural networks

Anisotropic Diffusion Descriptors

Learning a Predictable and Generative Vector Representation for Objects

Robust Nonrigid Registration by Convex Optimization

Dense Human Body Correspondences Using Convolutional Networks

Learning class‐specific descriptors for deformable shapes using localized spectral convolutional networks

The stitched puppet: A graphical model of 3D human shape and pose

Geodesic Convolutional Neural Networks on Riemannian Manifolds

FAUST: Dataset and Evaluation for 3D Mesh Registration

Dense Non-rigid Shape Correspondence Using Random Forests

3D ShapeNets: A deep representation for volumetric shapes

Hierarchical Framework for Shape Correspondence

Soft Maps Between Surfaces

Functional maps

Exploring collections of 3D models using fuzzy correspondences

The wave kernel signature: A quantum mechanical approach to shape analysis

Coarse‐to‐Fine Combinatorial Matching for Dense Isometric Shape Correspondence

Blended intrinsic maps

Unique Signatures of Histograms for Local Surface Description

One Point Isometric Matching with the Heat Kernel

Möbius voting for surface correspondence

Differential Representations for Mesh Processing

Learning a correlated model of identity and pose-dependent body shape variation for real-time synthesis

Efficient Computation of Isometry-Invariant Distances Between Surfaces

Generalized multidimensional scaling: A framework for isometry-invariant partial surface matching

A Theoretical and Computational Framework for Isometry Invariant Recognition of Point Cloud Data

SCAPE: shape completion and animation of people

The space of human body shapes: reconstruction and parameterization from range scans

Articulated body deformation from range scan data

Supplementary material (appendix) for the paper https://http://imagine.enpc.fr/~groueixt/ 3D-CODED/index.html

Numerical Geometry of Non-Rigid Shapes

Discrete Differential-Geometry Operators for Triangulated 2-Manifolds

Eurographics Symposium on Geometry Processing 2009 a Concise and Provably Informative Multi-scale Signature Based on Heat Diffusion

Field of Study

Venue Information

Name

Type

URL

Alternate Names