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

Photoelectrochemical Dehydrogenative Cross-Coupling of Aliphatic C-H Bonds via an In-Situ-Generated Hypervalent Iodine Mediator.

Organic letters·2026
Same author

Human papillomavirus16 E7 enhances cell stemness by regulating the APC2/SPIN4/β-catenin axis in cervical cancer.

Oncogenesis·2026
Same author

Radical docking-migration: a powerful strategy for difunctionalization of alkenes and alkynes.

Chemical science·2026
Same author

Hypoxic glycolysis-driven histone lactylation activates NHE7 to promote endometrial cancer progression via COX6C-mediated endoplasmic reticulum stress.

Apoptosis : an international journal on programmed cell death·2026
Same author

CSN5 overexpression promotes the integral progression of cervical cancer by enhancing ENO3-mediated glycolysis.

Apoptosis : an international journal on programmed cell death·2026
Same author

<i>Cis</i>-difluoromethyl hetarylative dearomatization by a radical docking-migration cascade.

Chemical science·2025

相关实验视频

Updated: May 7, 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

2.6K

伪标识指导像素对比用于域自适应语义细分的域自适应语义细分.

Jianzi Xiang1, Cailu Wan1, Zhu Cao2

  • 1The Key Laboratory of Smart Manufacturing in Energy Chemical Process, Ministry of Education, East China University of Science and Technology, Shanghai, 200237, China.

Scientific reports
|December 31, 2024
PubMed
概括

本研究引入了伪标签指导像素对比 (PGPC) 来改善无监督域适应语义细分. PGPC增强了功能多样性,导致更准确的图像理解,而无需昂贵的像素级注释.

关键词:
相反的学习学习.语义细分 语义细分是指语义细分.无监督的域名适应

更多相关视频

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

318
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

8.9K

相关实验视频

Last Updated: May 7, 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

2.6K
Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

318
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

8.9K

科学领域:

  • 计算机视觉 计算机视觉
  • 机器学习 机器学习
  • 人工智能的人工智能

背景情况:

  • 语义细分需要像素级的注释,这是昂贵和耗时的.
  • 无监督域调整 (UDA) 用于语义细分利用标记合成数据用于未标记的现实数据.
  • 使用对比学习的现有UDA方法忽视了类内特征多样性,导致预测错误.

研究的目的:

  • 为了解决UDA中目前对比式学习方法的局限性,用于语义细分.
  • 提出一个新的框架,伪标签指导像素对比 (PGPC),以提高类预测的准确性.
  • 从目标图像中有效利用信息,同时最大限度地减少伪标签噪音.

主要方法:

  • 开发了伪标签指导像素对比 (PGPC) 框架.
  • 纳入了考虑类内特征多样性的对比学习策略.
  • 研究了利用目标图像信息而不引入噪音的方法.

主要成果:

  • 在标准的UDA基准指标上,PGPC框架表现优越.
  • 使用DAFormer.实现了5.1%的mIoU (GTA5到Cityscapes) 和4.6%的mIoU (SYNTHIA到Cityscapes) 的相对改善.
  • 拟议的方法增强了现有的UDA方法,而不会增加模型的复杂性.

结论:

  • PGPC有效地克服了以前用于语义细分的UDA方法的局限性.
  • 该框架通过考虑类内的特征多样性来提高模型准确性.
  • 在语义细分任务中,PGPC为各种UDA方法提供了有价值的增强.