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

A Lightweight Zero-Trust-Based Authentication Framework for UAV Systems.

Sensors (Basel, Switzerland)·2026
Same author

Heterogeneous photocatalytic aldehyde alkylative amination for multicomponent synthesis of α-branched amines.

National science review·2026
Same author

Installation of d3-methyl group to drugs by continuous-flow solid-phase synthesis.

Nature communications·2025
Same author

The role of LncRNA SNHG3 in human cancers.

Discover oncology·2025
Same author

Comprehensive analysis reveals MCM4 as a biomarker for guiding therapies and immunomodulatory role in skin cutaneous melanoma.

Journal of Cancer·2025
Same author

GoFormer: A GoLPP inspired transformer for functional brain graph learning and classification.

Neural networks : the official journal of the International Neural Network Society·2025

相关实验视频

Updated: May 16, 2025

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

438

基于多重注意力机制的ID不敏感深度假冒检测模型.

Yuncan Sheng1, Zhengrui Zou2, Zongxuan Yu2

  • 1School of Information and Communication Engineering, Hainan University, Haikou, 570228, China.

Scientific reports
|April 1, 2025
PubMed
概括

这项研究引入了一种新的多重注意力深度假冒检测模型. 该模型增强了纹理特征,检测了多尺度的文物,并融合了本地和全球信息,以改进深度假冒识别.

关键词:
注意地图注意地图深度假冒的检测检测.多个尺度的文物检测检测.质地特征增强增强质地特征增强

更多相关视频

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
06:37

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023

2.5K
Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images
08:20

Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images

Published on: October 27, 2023

1.3K

相关实验视频

Last Updated: May 16, 2025

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

438
Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
06:37

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023

2.5K
Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images
08:20

Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images

Published on: October 27, 2023

1.3K

科学领域:

  • 计算机科学 计算机科学
  • 人工智能的人工智能
  • 数字法医学数字法医学

背景情况:

  • 深度假冒技术促进了被操纵的面部内容在线传播,造成了重大的社会风险.
  • 目前的深度假冒检测方法往往忽视了本地和全球图像特征之间的相互作用,并未能减轻身份泄露,导致性能不佳,特别是在跨数据集场景中.

研究的目的:

  • 开发一个强大的深度假冒检测模型,解决现有方法的局限性.
  • 提高深度假冒检测的准确性和通用性,特别是在交叉数据集评估中.

主要方法:

  • 提出了一个多重注意深度假冒检测模型,包括三个关键组件.
  • 纹理特征增强:利用CondenseNet来有效地提取纹理特征,保持细节.
  • 多尺度文物检测:纳入了一个模块,以识别各种尺度的操纵区域,最大限度地减少身份信息的影响.
  • 多重注意力机制:生成多个注意力地图来优先考虑图像区域,并为增强分类而进行接纹理和局部特征.

主要成果:

  • 该模型在面部操纵检测的FaceForensics++和DFDC基准测试中表现出卓越的性能.
  • 在Celeb-DF-v2的交叉数据集评估中取得了最先进的结果,表明了强大的通用性.

结论:

  • 拟议的多重注意模式有效地整合了本地和全球特征,以准确检测深度假冒.
  • 该方法对现实世界的应用程序具有显著的前景,需要对操纵的面部内容进行强有力的检测.