PartNet: A Large-Scale Benchmark for Fine-Grained and Hierarchical Part-Level 3D Object Understanding (2018-12-06T00:00:00.000000Z)

TL;DR

This work presents PartNet, a consistent, large-scale dataset of 3D objects annotated with fine-grained, instance-level, and hierarchical 3D part information, and proposes a baseline method for part instance segmentation that is superior performance over existing methods.

Abstract

We present PartNet: a consistent, large-scale dataset of 3D objects annotated with fine-grained, instance-level, and hierarchical 3D part information. Our dataset consists of 573,585 part instances over 26,671 3D models covering 24 object categories. This dataset enables and serves as a catalyst for many tasks such as shape analysis, dynamic 3D scene modeling and simulation, affordance analysis, and others. Using our dataset, we establish three benchmarking tasks for evaluating 3D part recognition: fine-grained semantic segmentation, hierarchical semantic segmentation, and instance segmentation. We benchmark four state-of-the-art 3D deep learning algorithms for fine-grained semantic segmentation and three baseline methods for hierarchical semantic segmentation. We also propose a baseline method for part instance segmentation and demonstrate its superior performance over existing methods.

Authors

Hao Su

10 papers

Kaichun Mo

5 papers

L. Guibas

28 papers

TL;DR

Abstract

Authors

References53 items

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

Learning to group and label fine-grained shape components

Physical Primitive Decomposition

VoxSegNet: Volumetric CNNs for Semantic Part Segmentation of 3D Shapes

SAGNet

Predictive and generative neural networks for object functionality

Monte Carlo convolution for learning on non-uniformly sampled point clouds

VirtualHome: Simulating Household Activities Via Programs

PointGrid: A Deep Network for 3D Shape Understanding

Deep Functional Dictionaries: Learning Consistent Semantic Structures on 3D Models from Functions

SpiderCNN: Deep Learning on Point Sets with Parameterized Convolutional Filters

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

SPLATNet: Sparse Lattice Networks for Point Cloud Processing

Robust Watertight Manifold Surface Generation Method for ShapeNet Models

Dynamic Graph CNN for Learning on Point Clouds

PointCNN: Convolution On X-Transformed Points

CHALET: Cornell House Agent Learning Environment

AI2-THOR: An Interactive 3D Environment for Visual AI

3D Semantic Segmentation with Submanifold Sparse Convolutional Networks

SGPN: Similarity Group Proposal Network for 3D Point Cloud Instance Segmentation

O-CNN

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

Visual Semantic Planning Using Deep Successor Representations

Deformation-driven shape correspondence via shape recognition

Learning hierarchical shape segmentation and labeling from online repositories

Semantic3D.net: A new Large-scale Point Cloud Classification Benchmark

Escape from Cells: Deep Kd-Networks for the Recognition of 3D Point Cloud Models

ScanNet: Richly-Annotated 3D Reconstructions of Indoor Scenes

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

SyncSpecCNN: Synchronized Spectral CNN for 3D Shape Segmentation

A scalable active framework for region annotation in 3D shape collections

Learning how objects function via co-analysis of interactions

ShapeNet: An Information-Rich 3D Model Repository

Recurrent Instance Segmentation

Shape2Pose

Parameter Learning and Convergent Inference for Dense Random Fields

Active co-analysis of a set of shapes

Co‐Segmentation of 3D Shapes via Subspace Clustering

Exploring Shape Variations by 3D‐Model Decomposition and Part‐based Recombination

Exploration of continuous variability in collections of 3D shapes

Learning 3D mesh segmentation and labeling

A benchmark for 3D mesh segmentation

A framework for the objective evaluation of segmentation algorithms using a ground-truth of human segmented 3D-models

Et al

Mesh Segmentation - A Comparative Study

Parts of recognition

The Hungarian method for the assignment problem

Linking WordNet to 3D Shapes

and L

Material for : Learning to Predict Part Mobility from a Single Static Snapshot

Visualizing Data using t-SNE

Template visualization (1/3). We present the templates for table and chair. The lamp template is shown in the main paper. The And-nodes are drawn in solid lines and Or-nodes in dash lines

Avg Bag Bed Bott Bowl Chair Clock Dish Disp Door Ear Fauc Hat Key Knife Lamp Lap Micro Mug Frid Scis Stora Table Trash Vase S1 72

Field of Study

Journal Information

Name

Page

Venue Information

Name

Type

URL

Alternate Names