Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

ChSCL9 negatively regulates citric acid accumulation via repressing PH4-PH5 module in kumquat.

Plant physiology·2026

Same author

Structural basis for the lack of immunogenicity of a Cryptosporidium octapeptide: anchor switching induces MHC-I groove remodeling and instability.

International journal of biological macromolecules·2026

Same author

Efficacy of repetitive transcranial magnetic stimulation combined with soft robotic glove training for slight-to-moderate post-stroke upper limb spasticity: a randomized controlled trial.

Topics in stroke rehabilitation·2026

Same author

Mechanism of Structure and Property Evolution of ABS During Multiple Extrusion and Aging Degree Prediction via Image Recognition Technology.

Polymers·2026

Same author

Topological Data Analysis in Materials Science: Principles, Machine Learning Integration, and Application Landscapes.

Chemical reviews·2026

Same author

Altered temporal variability-based functional reorganization of brain networks predicts motor outcome after stroke.

Journal of neuroengineering and rehabilitation·2026

Same journal

MesoSplats: Texture Synthesis with Gaussian Splatting.

IEEE transactions on visualization and computer graphics·2026

Same journal

GLLA: A Unified Force-Directed Graph Layout Framework Supporting Local Adjustments.

IEEE transactions on visualization and computer graphics·2026

Same journal

Multi-Perception Crowd: Learning to combine entity and implicit perception for diverse crowd simulation.

IEEE transactions on visualization and computer graphics·2026

Same journal

Hiding in Plain Sight: Camouflaging Real-world Objects.

IEEE transactions on visualization and computer graphics·2026

Same journal

RTF2Mesh: Restricted Tangent Face Based Mesh Compression With Neural Displacement Fields.

IEEE transactions on visualization and computer graphics·2026

Same journal

Practical Occluder Generation for Mobile Games.

IEEE transactions on visualization and computer graphics·2026

See all related articles

Search research articles

Related Experiment Video

Updated: May 24, 2025

From Voxels to Knowledge: A Practical Guide to the Segmentation of Complex Electron Microscopy 3D-Data

From Voxels to Knowledge: A Practical Guide to the Segmentation of Complex Electron Microscopy 3D-Data

Published on: August 13, 2014

Aligning Instance-Semantic Sparse Representation Towards Unsupervised Object Segmentation and Shape Abstraction With

Jiaxin Li, Hongxing Wang, Jiawei Tan

IEEE Transactions on Visualization and Computer Graphics

|March 3, 2025

Summary

This summary is machine-generated.

This study introduces a novel unsupervised framework for semantic-aware 3D shape representation. It achieves joint instance and semantic segmentation, enabling computers to understand object parts without costly annotations or multi-stage training.

More Related Videos

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

Related Experiment Videos

Last Updated: May 24, 2025

From Voxels to Knowledge: A Practical Guide to the Segmentation of Complex Electron Microscopy 3D-Data

From Voxels to Knowledge: A Practical Guide to the Segmentation of Complex Electron Microscopy 3D-Data

Published on: August 13, 2014

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

Area of Science:

Computer Vision
3D Shape Analysis
Machine Learning

Background:

Current 3D shape representation methods often rely on supervised learning with extensive annotations or unsupervised methods with strong semantic priors and multi-stage training.
These limitations hinder the generalization and deployment of shape understanding and reasoning systems.
There is a need for efficient, unsupervised methods that can accurately represent 3D object shapes using meaningful parts.

Purpose of the Study:

To develop a one-stage, fully unsupervised framework for semantic-aware 3D shape representation.
To enable joint instance segmentation, semantic segmentation, and shape abstraction of object parts.
To improve the interpretability and repeatability of 3D shape representations.

Main Methods:

A one-stage, fully unsupervised framework utilizing sparse representation and feature alignment in a high-dimensional space.
Sparse latent membership pursuit to model object part features as sparse convex combinations of point features.
An attention-based strategy for aligning instance- and semantic-level object part features, ensuring geometric reusability and semantic consistency.
Cascade unfrozen learning on geometric parameters for semantic disambiguation.

Main Results:

Achieved joint instance and semantic segmentation along with shape abstraction of object parts.
Demonstrated the ability to produce repeatable primitives for 3D shape representation.
Provided coherent semantic interpretations of 3D object shapes across categories.
Validated performance on benchmark datasets, confirming unsupervised, one-stage operation without annotations or semantic priors.

Conclusions:

The proposed framework offers a significant advancement in unsupervised 3D shape representation.
It effectively addresses the limitations of existing supervised and unsupervised methods.
The approach facilitates more robust and generalizable 3D shape understanding and reasoning.