Unsupervised 3D Learning for Shape Analysis via Multiresolution Instance Discrimination (2020-08-03T00:00:00.000000Z)

TL;DR

An unsupervised method for learning a generic and efficient shape encoding network for different shape analysis tasks and surpasses existing supervised methods by a large margin for fine-grained shape segmentation on the PartNet dataset.

Abstract

We propose an unsupervised method for learning a generic and efficient shape encoding network for different shape analysis tasks. Our key idea is to jointly encode and learn shape and point features from unlabeled 3D point clouds. For this purpose, we adapt HRNet to octree-based convolutional neural networks for jointly encoding shape and point features with fused multiresolution subnetworks and design a simple-yet-efficient Multiresolution Instance Discrimination (MID) loss for jointly learning the shape and point features. Our network takes a 3D point cloud as input and output both shape and point features. After training, Our network is concatenated with simple task-specific back-ends and fine-tuned for different shape analysis tasks. We evaluate the efficacy and generality of our method with a set of shape analysis tasks, including shape classification, semantic shape segmentation, as well as shape registration tasks. With simple back-ends, our network demonstrates the best performance among all unsupervised methods and achieves competitive performance to supervised methods. For fine-grained shape segmentation on the PartNet dataset, our method even surpasses existing supervised methods by a large margin.

Authors

Zhirong Wu

2 papers

Yang Liu

2 papers

Xin Tong

2 papers

TL;DR

Abstract

Authors

References71 items

ScaleNet: An Unsupervised Representation Learning Method for Limited Information

Generative Adversarial Networks

PointContrast: Unsupervised Pre-training for 3D Point Cloud Understanding

Mapping in a cycle: Sinkhorn regularized unsupervised learning for point cloud shapes

Momentum Contrast for Unsupervised Visual Representation Learning

Unsupervised Multi-Task Feature Learning on Point Clouds

Unsupervised Feature Learning for Point Cloud Understanding by Contrasting and Clustering Using Graph Convolutional Neural Networks

Deep High-Resolution Representation Learning for Visual Recognition

Deep Closest Point: Learning Representations for Point Cloud Registration

Unsupervised Detection of Distinctive Regions on 3D Shapes

KPConv: Flexible and Deformable Convolution for Point Clouds

Relation-Shape Convolutional Neural Network for Point Cloud Analysis

PointNetLK: Robust & Efficient Point Cloud Registration Using PointNet

PartNet: A Recursive Part Decomposition Network for Fine-Grained and Hierarchical Shape Segmentation

Self-Supervised Deep Learning on Point Clouds by Reconstructing Space

3D Shape Analysis: Fundamentals, Theory, and Applications

3D Point Capsule Networks

PartNet: A Large-Scale Benchmark for Fine-Grained and Hierarchical Part-Level 3D Object Understanding

PointConv: Deep Convolutional Networks on 3D Point Clouds

Transductive Centroid Projection for Semi-supervised Large-Scale Recognition

PPF-FoldNet: Unsupervised Learning of Rotation Invariant 3D Local Descriptors

Unsupervised Feature Learning via Non-parametric Instance Discrimination

SpiderCNN: Deep Learning on Point Sets with Parameterized Convolutional Filters

Point convolutional neural networks by extension operators

SO-Net: Self-Organizing Network for Point Cloud Analysis

SPLATNet: Sparse Lattice Networks for Point Cloud Processing

A Papier-Mache Approach to Learning 3D Surface Generation

CosFace: Large Margin Cosine Loss for Deep Face Recognition

Dynamic Graph CNN for Learning on Point Clouds

PointCNN: Convolution On X-Transformed Points

FoldingNet: Point Cloud Auto-Encoder via Deep Grid Deformation

3D Semantic Segmentation with Submanifold Sparse Convolutional Networks

Dynamic Routing Between Capsules

Large-Scale 3D Shape Reconstruction and Segmentation from ShapeNet Core55

O-CNN

Learning Representations and Generative Models for 3D Point Clouds

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

Learning Features by Watching Objects Move

3D Shape Segmentation with Projective Convolutional Networks

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

SyncSpecCNN: Synchronized Spectral CNN for 3D Shape Segmentation

Geometric Deep Learning on Graphs and Manifolds Using Mixture Model CNNs

OctNet: Learning Deep 3D Representations at High Resolutions

Learning a Probabilistic Latent Space of Object Shapes via 3D Generative-Adversarial Modeling

Fast Global Registration

Temporal Ensembling for Semi-Supervised Learning

Learning shape correspondence with anisotropic convolutional neural networks

Go-ICP: A Globally Optimal Solution to 3D ICP Point-Set Registration

3D-R2N2: A Unified Approach for Single and Multi-view 3D Object Reconstruction

Colorful Image Colorization

Unsupervised 3D shape segmentation and co-segmentation via deep learning

Deep Residual Learning for Image Recognition

ShapeNet: An Information-Rich 3D Model Repository

VoxNet: A 3D Convolutional Neural Network for real-time object recognition

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

Unsupervised Visual Representation Learning by Context Prediction

U-Net: Convolutional Networks for Biomedical Image Segmentation

Multi-view Convolutional Neural Networks for 3D Shape Recognition

Sparse 3D convolutional neural networks

An Analysis of Unsupervised Pre-training in Light of Recent Advances

FAUST: Dataset and Evaluation for 3D Mesh Registration

Representation Learning: A Review and New Perspectives

Why Does Unsupervised Pre-training Help Deep Learning?

Dimensionality Reduction by Learning an Invariant Mapping

BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding

MeshCNN: A network with an edge

ModelNet benchmark leaderboard

Improving Language Understanding by Generative Pre-Training

Deep Learning

3 D ShapeNets : A Deep Representation for Volumetric Shape Modeling

Visualizing Data using t-SNE

Field of Study

Venue Information

Name

Type

URL