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

Numerical Simulation and Theoretical Analysis of Flexural Strengthening of Undamaged RC Beams with Steel Strand Mesh-Reinforced ECC.

Materials (Basel, Switzerland)·2026
Same author

SinColor: Uncertainty-Guided Single-Step Diffusion for Image Colorization.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
Same author

Invited review: Manufacturing Whey Protein Colloidal Particles via Liquid Antisolvent Precipitation Method: Particle Formation Mechanism and Ingredient Functionality Aspects.

Journal of dairy science·2026
Same author

Integrating SAM Supervision for 3D Weakly Supervised Point Cloud Segmentation.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
Same author

For Intermediate-Size Sessile Serrated Lesions, Durable Clearance Should Remain the Decisive Endpoint.

United European gastroenterology journal·2026
Same author

Therapeutic potential of wogonoside in hypertension-induced cardiac injury: Targeting apoptosis and MAPK signaling pathway.

The Journal of nutritional biochemistry·2026

相关实验视频

Updated: May 24, 2025

Measuring the Structure, Composition, and Change of Underwater Environments with Large-area Imaging
09:19

Measuring the Structure, Composition, and Change of Underwater Environments with Large-area Imaging

Published on: April 18, 2025

286

在户外4D点云上的弱监督细分与渐进的4D分组.

Hanyu Shi, Fayao Liu, Zhonghua Wu

    IEEE transactions on pattern analysis and machine intelligence
    |March 3, 2025
    PubMed
    概括

    本研究介绍了使用最小注释进行3D点云细分的渐进4D分组方法. 这种方法产生了高质量的伪标签,显著优于SemanticKITTI上以前的方法,仅有0.001%的数据.

    科学领域:

    • 计算机视觉 计算机视觉
    • 机器学习 机器学习
    • 3D数据分析 3D数据分析

    背景情况:

    • 弱监督的3D点云细分方法旨在减少注释工作.
    • 之前的工作 (W4DTS) 用了0.001%的分点进行细分,但遭受了低质量的伪标签.
    • 极其有限的注释对有效的伪标签生成构成了挑战.

    研究的目的:

    • 开发一种方法来生成高质量的伪标签,在3D点云细分中非常稀疏的注释.
    • 在极端注释约束下,改进弱监督的细分模型的性能.
    • 通过额外的学习策略来增强渐进的4D分组方法.

    主要方法:

    • 提出了一种渐进的4D分组方法,从稀疏的注释和未注释点汇总空间和时间信息.
    • 引入了跨框架对比学习,以改进跨不同时间框架的特征表示.
    • 实施了本地一致性学习,以在细分过程中执行邻里协议.

    主要成果:

    • 在仅使用0.001%注释的SemanticKITTI数据集上实现了显著的性能改进,超过了之前的最佳方法.
    • 在SemanticPOSS和ScribbleKITTI数据集上展示了竞争性结果,接近完全监督模型的性能.
    • 验证了渐进4D分组,跨框架对比学习和局部一致性学习的有效性.

    更多相关视频

    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
    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 24, 2025

    Measuring the Structure, Composition, and Change of Underwater Environments with Large-area Imaging
    09:19

    Measuring the Structure, Composition, and Change of Underwater Environments with Large-area Imaging

    Published on: April 18, 2025

    286
    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
    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

    结论:

    • 提出的渐进4D分组方法有效地产生高质量的伪标签,即使是非常稀疏的注释.
    • 这种方法在严重的注释预算限制下,显著提升了监督较弱的3D点云细分.
    • 该框架为在现实应用中高效的3D数据注释和细分提供了一个有希望的方向.