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

Search research articles

Related Experiment Video

Updated: Feb 23, 2026

Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches

Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches

Published on: December 15, 2023

Analyzing the Training Processes of Deep Generative Models.

Mengchen Liu, Jiaxin Shi, Kelei Cao

IEEE Transactions on Visualization and Computer Graphics

|September 4, 2017

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

Layer-specific facial soft-tissue thickness in 1174 Chinese adults: Implications for finite-element headforms and ergonomic design.

Ergonomics·2026

Same author

Ribavirin mitigates Alzheimer's disease model phenotypes by enhancing lysosomal function and suppressing ISR.

Bioorganic chemistry·2026

Same author

[Mechanism on protective effect of acupuncture combined with medication pretreatment for acute gastric ulcer rats based on metabolomics].

Zhongguo zhen jiu = Chinese acupuncture & moxibustion·2026

Same author

A light-controlled hydrogel for regulating oxidative stress in the treatment of psoriasis.

International journal of pharmaceutics·2026

Same author

Manganese-based activated carbon composites promote nitrogen removal in low temperature constructed wetlands via enhanced extracellular electron transfer.

Environmental research·2026

Same author

Fabrication and characterization of edible composite films derived from xylanase-enzymatic flaxseed gum and citric acid-modified soybean protein isolate: Intermolecular interactions and functional efficacy.

International journal of biological macromolecules·2026

Same journal

FGO-SLAM++: Real-time Geometry-Aware Gaussian SLAM with Continuous Opacity Field.

IEEE transactions on visualization and computer graphics·2026

Same journal

Blue Noise Dithering for Reservoir-based Spatio-temporal Importance Resampling.

IEEE transactions on visualization and computer graphics·2026

Same journal

ROS-GS: Relightable Outdoor Scenes With Gaussian Splatting.

IEEE transactions on visualization and computer graphics·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

See all related articles

Training deep generative models (DGMs) is complex. This study introduces a visual analytics approach to understand and diagnose DGM training, aiding machine learning experts in analyzing complex deep learning models.

Area of Science:

Artificial Intelligence
Machine Learning
Data Visualization

Background:

Deep generative models (DGMs) are powerful tools for unsupervised and semi-supervised learning.
Training DGMs presents significant challenges compared to other deep models like convolutional neural networks (CNNs).
Effective diagnosis of DGM training failures is crucial for model development.

Purpose of the Study:

To develop a visual analytics approach for enhanced understanding and diagnosis of deep generative model training.
To provide tools for machine learning experts to interpret complex training dynamics.
To identify the root causes of training failures in DGMs.

Main Methods:

Extraction of extensive time series data representing DGM training dynamics.

Related Experiment Videos

Last Updated: Feb 23, 2026

Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches

Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches

Published on: December 15, 2023

Implementation of a blue-noise polyline sampling scheme to manage visual clutter while preserving critical data.

Development of a credit assignment algorithm to trace neuron contributions to training failures.

Main Results:

The visual analytics approach effectively aids in understanding overall DGM training processes.
The credit assignment algorithm successfully identifies neuron contributions to training anomalies.
Case studies with experts validate the approach's utility in diagnosing DGM training.

Conclusions:

The proposed visual analytics method improves the interpretability and diagnosis of deep generative model training.
The approach is adaptable for analyzing other deep learning models, including CNNs.
This work offers valuable insights for researchers and practitioners working with complex deep learning architectures.