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

相关实验视频

Updated: May 22, 2025

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

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

8.9K

ADAMT:适应式分布式多任务学习,用于移动临时网络中高效的图像识别.

Jia Zhao1, Wei Zhao2, Yunan Zhai3

  • 1School of Computer Technology and Engineering, Changchun Institute of Technology, Changchun, China; College of Artificial Intelligence Technology, Changchun Institute of Technology, Changchun, China; School of Electronics Engineering and Computer Science, Peking University, Beijing, China.

Neural networks : the official journal of the International Neural Network Society
|March 12, 2025
PubMed
概括

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

FAM91A1-TBC1D23 complex structure reveals human genetic variations susceptible for PCH.

Proceedings of the National Academy of Sciences of the United States of America·2023
Same author

Editorial: Natural products in the treatment of neurological diseases: identification of novel active compounds and therapeutic targets.

Frontiers in pharmacology·2023
Same author

Higher visceral adipose tissue is associated with decreased memory suppression ability on food-related thoughts: A 1-year prospective ERP study.

Appetite·2023
Same author

Novel role for epalrestat: protecting against NLRP3 inflammasome-driven NASH by targeting aldose reductase.

Journal of translational medicine·2023
Same author

Electroencephalographic oscillations of alpha and beta rhythms during phrase-guessing procedure.

Cognitive neurodynamics·2023
Same author

Integrative analysis of metabolome, proteome, and transcriptome for identifying genes influencing total lignin content in <i>Populus trichocarpa</i>.

Frontiers in plant science·2023

我们开发了ADAMT,这是一个适应式分布式多任务学习框架,用于移动特设网络. 它通过优化本地培训和协作更新来提高资源有限的环境中的图像识别效率.

科学领域:

  • 计算机科学 计算机科学
  • 人工智能的人工智能
  • 移动计算 移动计算

背景情况:

  • 在移动特设网络中分布式机器学习面临来自有限资源,非IID数据和动态拓学的挑战.
  • 现有的方法往往需要集中协调和稳定的网络,这限制了实际应用.
  • 高效的图像识别对于移动环境中的边缘设备至关重要.

研究的目的:

  • 提出ADAMT,一个适应式分布式多任务学习框架,用于在资源有限的移动特设网络中高效的图像识别.
  • 在动态和资源有限的环境中解决现有的分布式学习方法的局限性.
  • 为了实现个性化的模型训练和边缘设备上的协作参数更新.

主要方法:

  • ADAMT使用特征扩展机制来增强本地模型的表达力.
  • 一种深度散列技术促进了高效的设备上检索和多任务融合.
  • 适应式通信策略根据网络条件和节点可靠性动态调整模型更新.

主要成果:

  • 在ImageNet数据集上,ADAMT实现了0.867的top-1精度,超过了最先进的方法.
  • 与分布式SGD相比,该框架显著降低了通讯开支,并加速了汇聚的2.69倍.
  • 适应性沟通策略有效地平衡了模型性能和资源消耗.
关键词:
分散式的学习学习是分散式的分布式多任务学习图像识别 图像识别 图像识别移动定时网络是移动定时网络.

更多相关视频

A Methodology for Capturing Joint Visual Attention Using Mobile Eye-Trackers
12:39

A Methodology for Capturing Joint Visual Attention Using Mobile Eye-Trackers

Published on: January 18, 2020

7.5K
Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

343

相关实验视频

Last Updated: May 22, 2025

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

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

8.9K
A Methodology for Capturing Joint Visual Attention Using Mobile Eye-Trackers
12:39

A Methodology for Capturing Joint Visual Attention Using Mobile Eye-Trackers

Published on: January 18, 2020

7.5K
Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

343

结论:

  • ADAMT为移动特设网络中的分布式学习提供了一种高效,强大的解决方案.
  • 该框架非常适合资源有限的环境,在边缘设备上实现先进的机器学习.
  • 这项工作推动了分布式学习算法的设计,用于实际的边缘人工智能应用.