Search research articles

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 Journal Methods Collections JoVE Encyclopedia of Experiments 存档

教育

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual 教师资源中心教师网站

使用条款与条件

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

Serum FAP⁺ exosome-based flow cytometric detection: potential utility in early diagnosis, disease progression assessment, and surgical response monitoring of lung adenocarcinoma.

Journal of nanobiotechnology·2026

Same author

RNF145 promotes inflammation in recurrent spontaneous abortion through inducing the K63-linked polyubiquitination of IRF3.

Journal of reproductive immunology·2026

Same author

Artesunate reverses gefitinib resistance in lung adenocarcinoma by inducing ferroptosis and suppressing the Wnt/β-catenin pathway.

European journal of pharmacology·2026

Same author

ROFI: a deep learning-based ophthalmic sign-preserving and reversible patient face anonymizer.

NPJ digital medicine·2025

Same author

GPRC5A+ myCAFs promote ESCC progression via TGF-β-induced fibroblast activation and ANXA1-mediated M2 macrophage polarization.

International immunopharmacology·2025

Same author

InstantGroup: Instant Template Generation for Scalable Group of Brain MRI Registration.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2025

Search research articles

相关实验视频

Updated: Jun 18, 2025

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

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

Published on: November 30, 2022

EfficientQ:一种高效准确的训练后神经网络量化方法,用于医疗图像细分.

Rongzhao Zhang¹, Albert C S Chung¹

¹Department of Computer Science and Engineering, The Hong Kong University of Science and Technology, Hong Kong, China.

Medical image analysis

|August 2, 2024

概括

此摘要是机器生成的。

EfficientQ使深度神经网络的快速准确的训练后定量化成为可能. 这种方法显著减少了模型量化所需的时间和数据,使人工智能更容易获得.

关键词:

深度学习是一种深度学习.图像细分图像细分图像细分模型加速的加速模型.模型的压缩压缩.神经网络的量化神经网络量化.培训后的量化定量化

更多相关视频

Automated Image-Based Quantification of Neutrophil Extracellular Traps Using NETQUANT

Automated Image-Based Quantification of Neutrophil Extracellular Traps Using NETQUANT

Published on: November 27, 2019

Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application

Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application

Published on: April 14, 2023

相关实验视频

Last Updated: Jun 18, 2025

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

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

Published on: November 30, 2022

Automated Image-Based Quantification of Neutrophil Extracellular Traps Using NETQUANT

Automated Image-Based Quantification of Neutrophil Extracellular Traps Using NETQUANT

Published on: November 27, 2019

Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application

Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application

Published on: April 14, 2023

科学领域:

人工智能的人工智能
深度学习 (Deep Learning) 是一种深度学习.
计算机视觉计算机视觉

背景情况:

模型量子化通过减少比特宽度来压缩和加速深度神经网络.
当前的量子化方法在计算上可能很昂贵,需要对大型数据集进行大量微调.

研究的目的:

开发一种高效准确的训练后量化方法,命名为EfficientQ.
为了减少模型量化所需的时间和数据.

主要方法:

采用层级优化策略和乘数的交替方向方法 (ADMM) 实现快速融合.
包含重量规范化和自我适应注意力机制,以解决离散重量优化和类不平衡.

主要成果:

与医疗图像细分数据集 (LiTS,BraTS2020) 上现有的培训后量化方法相比,显示出更高的准确性和优化速度.
使用单个GPU和一个数据样本,在不到5分钟的时间内量子化一个3D UNet模型.

结论:

EfficientQ在培训后的量化效率和有效性方面提供了显著的改进.
该方法对人工智能的实际应用具有前景,特别是在资源有限的环境中.