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

Burn Injuries01:22

Burn Injuries

2.5K
Burn injuries occur when the skin and underlying tissues are damaged due to exposure to heat, electricity, chemicals, radiation, or friction. They can vary in severity, from minor superficial burns to severe deep burns that can be life-threatening.
The damage results in the death of skin cells, which can lead to a massive loss of fluid. Dehydration, electrolyte imbalance, and renal and circulatory failure follow, which can be fatal. Burn patients are treated with intravenous fluids to offset...
2.5K

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

[ATM/H2AX and repair of sperm-DNA damage during cryopreservation].

Zhonghua nan ke xue = National journal of andrology·2011
Same author

Predicting accident frequency at their severity levels and its application in site ranking using a two-stage mixed multivariate model.

Accident; analysis and prevention·2011
Same author

Photothermally enhanced photodynamic therapy delivered by nano-graphene oxide.

ACS nano·2011
Same author

[Characteristics of soil respiration in Phyllostachys edulis forest in Wanmulin Natural Reserve and related affecting factors].

Ying yong sheng tai xue bao = The journal of applied ecology·2011
Same author

Quality changes in sea urchin (Strongylocentrotus nudus) during storage in artificial seawater saturated with oxygen, nitrogen and air.

Journal of the science of food and agriculture·2011
Same author

Global effect of an RNA polymerase β-subunit mutation on gene expression in the radiation-resistant bacterium Deinococcus radiodurans.

Science China. Life sciences·2011

相关实验视频

Updated: Jun 27, 2025

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

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

Published on: November 30, 2022

2.7K

基于深度学习的燃烧图像细分的对抗性攻击和对抗性训练.

Luying Chen1, Jiakai Liang1, Chao Wang1

  • 1Zhejiang Integrated Circuits and Intelligent Hardware Collaborative Innovation Center, Hangzhou Dianzi University, Hangzhou, 317300, China.

Medical & biological engineering & computing
|May 1, 2024
PubMed
概括

这项研究引入了一种新的对抗式训练方法,以改进燃烧图像细分的深度学习模型,提高对自然主义干扰的准确性并减少训练时间.

关键词:
敌对的攻击是敌对的攻击.对抗性的训练是对抗性的训练.燃烧图像 燃烧图像深度学习是一种深度学习.图像细分的图像细分.

更多相关视频

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

524
Deep Neural Networks for Image-Based Dietary Assessment
13:19

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

9.1K

相关实验视频

Last Updated: Jun 27, 2025

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

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

Published on: November 30, 2022

2.7K
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

524
Deep Neural Networks for Image-Based Dietary Assessment
13:19

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

9.1K

科学领域:

  • 医疗成像医学成像
  • 计算机视觉 计算机视觉
  • 深度学习 (Deep Learning) 是一种深度学习.

背景情况:

  • 深度学习模型容易受到对抗性攻击,特别是在医学图像细分方面.
  • 医疗图像通常含有由于成像条件造成的噪音,影响模型性能.
  • 现有的对抗性训练方法可能无法完全解决烧伤图像中具有物理意义的干扰.

研究的目的:

  • 为燃烧图像细分提出一个对抗性训练方法,模拟自然现象启发的攻击.
  • 提高深度学习模型在分段烧伤图像中的稳定性和准确性.
  • 为了降低与对抗训练相关的计算成本.

主要方法:

  • 模拟的敌对攻击灵感来自自然现象.
  • 开发了一种专门的对抗性训练方法,用于燃烧图像细分.
  • 在专门的燃烧图像数据集上测试了该方法.
  • 进行了废弃实验,以验证单个损失组件.

主要成果:

  • 实现了对抗样本的82.19%的细分精度,比最初的54%有所增加.
  • 与传统的对抗训练方法相比,证明了1.97%的改进.
  • 与标准方法相比,训练时间显著减少.
  • 验证了拟议损失的有效性,并比较了不同对抗样本的性能.

结论:

  • 拟议的对抗式培训方法有效地提高了用于燃烧图像细分的深度学习模型性能.
  • 该方法增强了对自然主义图像干扰的稳定性,同时在计算上是高效的.
  • 这种技术在具有挑战性的条件下为可靠的医学图像分析提供了有希望的方向.