Jove
Visualize
联系我们
JoVE
x logofacebook logolinkedin logoyoutube logo
关于 JoVE
概览领导团队博客JoVE 帮助中心
作者
出版流程编辑委员会范围与政策同行评审常见问题投稿
图书馆员
用户评价订阅访问资源图书馆顾问委员会常见问题
研究
JoVE JournalMethods CollectionsJoVE Encyclopedia of Experiments存档
教育
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab Manual教师资源中心教师网站
使用条款与条件
隐私政策
政策

相关概念视频

您也可能阅读

相关文章

通过共同作者、期刊和引用图与本文相关的文章。

排序
Same author

Hydrogel-based lineage-specific gene therapy prevents recurrence of corticotroph tumors.

Molecular therapy. Oncology·2026
Same author

Dynamic outpatient sequencing on single examination equipment considering late arrivals.

Health care management science·2026
Same author

Illuminating LLM Coding Agents: Visual Analytics for Deeper Understanding and Enhancement.

IEEE transactions on visualization and computer graphics·2026
Same author

Interpretable SVM Model for Predicting CMV Infection in Seropositive Kidney Transplant Recipients: A Single-Center Retrospective Study.

Infection and drug resistance·2026
Same author

High-Breakdown, Thermally Stable Ceramic Fiber-Reinforced Aramid/Polyphenylene Sulfide Composite Papers for High-Power Electrical Applications.

ACS applied materials & interfaces·2026
Same author

A glucose-containing polysaccharide from Lycii fructus: Structural characterization and immunomodulatory activities in vitro and in vivo.

International journal of biological macromolecules·2026

相关实验视频

Updated: Jun 28, 2025

Combining Eye-tracking Data with an Analysis of Video Content from Free-viewing a Video of a Walk in an Urban Park Environment
08:25

Combining Eye-tracking Data with an Analysis of Video Content from Free-viewing a Video of a Walk in an Urban Park Environment

Published on: May 7, 2019

9.0K

视觉分析用于高效的图像探索和用户指导的图像标题.

Yiran Li, Junpeng Wang, Prince Aboagye

    IEEE transactions on visualization and computer graphics
    |April 16, 2024
    PubMed
    概括

    本研究介绍了一种用于探索大型图像数据集和评估图像字幕的视觉分析系统. 它有助于识别数据偏差,并通过交互式探索改进语言图像模型标题.

    科学领域:

    • 计算机视觉 计算机视觉
    • 数据科学数据科学数据科学
    • 人与计算机的交互

    背景情况:

    • 预先训练的语言图像模型提供了先进的视觉理解能力.
    • 视觉分析在探索大型图像数据集和评估标题质量方面面临着挑战.
    • 识别和减轻图像数据集中的数据偏差对于可靠的AI至关重要.

    研究的目的:

    • 开发一个视觉分析系统,以有效地探索大规模的图像数据集.
    • 为了使图像集合中的数据偏差能够被识别和理解.
    • 评估和指导语言图像模型的标题生成过程.

    主要方法:

    • 通过语言图像模型生成的标题的视觉检查,用于数据集偏差检测.
    • 分析视觉特征和文字标题之间的关联,以揭示模型的弱点.
    • 开发一个交互式界面来控制图片标题的生成.
    • 将视觉和文本分析集成到一个协调的系统中.

    主要成果:

    • 该系统可以更深入地了解视觉内容,并揭示根深蒂固的数据偏见.
    • 预先训练的语言图像模型的标题功能存在缺陷.
    • 交互式界面有效地指导了标题生成过程.

    更多相关视频

    Author Spotlight: Insights into Visual Cortex Research Through Wide-View fMRI Mapping
    07:11

    Author Spotlight: Insights into Visual Cortex Research Through Wide-View fMRI Mapping

    Published on: December 8, 2023

    1.5K
    Facilitating the Analysis of Immunological Data with Visual Analytic Techniques
    10:58

    Facilitating the Analysis of Immunological Data with Visual Analytic Techniques

    Published on: January 2, 2011

    10.1K

    相关实验视频

    Last Updated: Jun 28, 2025

    Combining Eye-tracking Data with an Analysis of Video Content from Free-viewing a Video of a Walk in an Urban Park Environment
    08:25

    Combining Eye-tracking Data with an Analysis of Video Content from Free-viewing a Video of a Walk in an Urban Park Environment

    Published on: May 7, 2019

    9.0K
    Author Spotlight: Insights into Visual Cortex Research Through Wide-View fMRI Mapping
    07:11

    Author Spotlight: Insights into Visual Cortex Research Through Wide-View fMRI Mapping

    Published on: December 8, 2023

    1.5K
    Facilitating the Analysis of Immunological Data with Visual Analytic Techniques
    10:58

    Facilitating the Analysis of Immunological Data with Visual Analytic Techniques

    Published on: January 2, 2011

    10.1K
  • 该系统促进视觉和文本数据之间的相互丰富.

    • 结论:

    • 开发的视觉分析系统有效地探索大型图像数据集和评估图像字幕.
    • 该系统有助于识别数据偏差并改善语言图像模型性能.
    • 领域从业者通过案例研究验证了系统的有效性.