Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Unsupervised visual learning of three-dimensional objects using a modular network architecture.

H Ando1, S Suzuki, T Fujita

  • 1ATR Human Information Processing Research Laboratories, Soraku-gun, Kyoto, Japan

Neural Networks : the Official Journal of the International Neural Network Society
|March 29, 2003
PubMed
Summary
This summary is machine-generated.

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

Comprehensive validation of early diagnostic algorithms for myocardial infarction in the emergency department.

QJM : monthly journal of the Association of Physicians·2023
Same author

Psychological stress during in vitro fertilization and embryo transfer is influenced by the patients' background and gender.

Reproductive medicine and biology·2018
Same author

Myogenic Heartbeat in the Primitive Crustacean Triops longicaudatus.

The Biological bulletin·2017
Same author

Wireless Multichannel Neural Recording With a 128-Mbps UWB Transmitter for an Implantable Brain-Machine Interfaces.

IEEE transactions on biomedical circuits and systems·2016
Same author

Multichannel neural recording with a 128 Mbps UWB wireless transmitter for implantable brain-machine interfaces.

Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual International Conference·2016
Same author

Synthesis of DNA fragments containing 2'-deoxy-4'-selenonucleoside units using DNA polymerases: comparison of dNTPs with O, S and Se at the 4'-position in replication.

Organic & biomolecular chemistry·2015
Same journal

Anchor-based disentanglement framework for incremental multi-view clustering.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

Complex-valued amplitude-phase interference modeling for adversarially robust classification.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

TraNce: Type-aware hypergraph neural network with biological mediators for drug repositioning.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

Decentralized ADMM for factorization-based Low-rank matrix estimation.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

Memristive neuromorphic circuit design inspired by the neural mechanisms of conditioned fear.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

Q-learning based asynchronous Boolean control networks stabilization with data loss.

Neural networks : the official journal of the International Neural Network Society·2026
See all related articles

This study introduces a novel modular network for clustering 3D object views using non-linear autoencoders. The approach effectively captures complex shape distributions, outperforming traditional clustering methods.

Area of Science:

  • Computer Vision
  • Machine Learning
  • Artificial Intelligence

Background:

  • Clustering multiple views of 3D objects is challenging due to complex shape distributions.
  • Existing methods like K-means and Gaussian Mixture Models struggle with non-linear data representations.

Purpose of the Study:

  • To present a modular network architecture for unsupervised clustering of multi-view 3D object data.
  • To demonstrate the capability of capturing multiple non-linear subspaces for improved shape representation.

Main Methods:

  • A novel network model based on a mixture of non-linear autoencoders is proposed.
  • An unsupervised training algorithm is formulated using maximum-likelihood estimation.
  • The model competes to encode multiple views of each 3D object.

Related Experiment Videos

Main Results:

  • The modular network effectively captures complex non-linear subspaces of view distributions.
  • Experimental results show superior performance compared to K-means and Gaussian Mixture Models.
  • Evaluation was conducted on synthetic wire-frame objects and real 3D object images.

Conclusions:

  • The proposed modular network architecture offers a more robust solution for 3D object clustering.
  • This approach advances the ability to represent and cluster complex 3D object shapes from multiple views.
  • The method shows significant potential for applications in computer vision and 3D data analysis.