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

Prediction of Clinical Outcomes Based on Phenotypic Grouping Using Left Ventricular and Left Atrial Strains in Patients with HFpEF.

European heart journal. Cardiovascular Imaging·2026
Same author

Multitask Artificial Intelligence-Based Electrocardiogram Tool for Preoperative Cardiac Testing in Noncardiac Surgery: Retrospective Cohort Study of Health Care Utilization and Costs.

Journal of medical Internet research·2026
Same author

Fully automated artificial intelligence-based echocardiographic analysis substantially reduces workflow time while preserving measurement accuracy: a pilot study.

Journal of cardiovascular imaging·2026
Same author

Point tracking as a temporal Cue for robust myocardial segmentation in echocardiography videos.

International journal of computer assisted radiology and surgery·2026
Same author

Validation of AI-enhanced ECG image analysis for identifying extreme cardiac magnetic resonance metrics in a cross-ethnic UK biobank study.

Scientific reports·2026
Same author

Machine-learning-derived phenotypes of hypertensive patients using multidimensional clinical and echocardiographic data including strain imaging.

European heart journal. Digital health·2026

相关实验视频

Updated: Jul 9, 2025

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

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

406

基于3D卷积神经网络的异常检测,结合卷积块注意模块,使用合并合.

In-Chang Hwang1, Hyun-Soo Kang1

  • 1Department of Information and Communication Engineering, School of Electrical and Computer Engineering, Chungbuk National University, Cheongju-si 28644, Republic of Korea.

Sensors (Basel, Switzerland)
|December 9, 2023
PubMed
概括
此摘要是机器生成的。

这项研究引入了一种新的深度学习模型,用于检测视频中的暴力行为,其性能优于现有的方法. 该方法有效地解决了用于实时视频监控的异常检测数据集的挑战.

关键词:
三维卷积的3D卷积在UBI战斗中,UBI战斗.在曲线下面的面积.计算机视觉 计算机视觉卷积块注意力模块的注意力模块同等错误率的错误率.视频监控视频监控视频监控

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

558
Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
06:37

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023

3.8K

相关实验视频

Last Updated: Jul 9, 2025

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

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

406
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

558
Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
06:37

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023

3.8K

科学领域:

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

背景情况:

  • 暴力犯罪的增加需要先进的监视能力.
  • 闭路电视 (CCTV) 系统对于实时情况分析至关重要.
  • 在视频中检测异常面临着由于数据集不平衡 (正常与异常事件) 的挑战.

研究的目的:

  • 开发一种有效的异常检测模型,用于识别视频中的暴力行为.
  • 为了应对视频数据集中有限的异常事件数据的挑战.
  • 为了改善公共安全的实时视频监控系统.

主要方法:

  • 异常检测被定义为一个二进制分类问题,使用数据集,如UBI-Fights,RWF-2000,UCSD Ped1/Ped2.
  • 视频被重建成图像补丁 (例如,3x3,4x4),模仿光场和视觉变压器输入.
  • 使用了一个3D卷积残余神经网络 (ResNet) 与卷积块注意模块 (CBAM).

主要成果:

  • 与现有方法相比,拟议的模型在检测异常行为方面表现出卓越的性能.
  • 在UBI-Fights数据集上,该模型实现了高精度 (0.9933),低损失 (0.0010) 和优秀的AUC (0.9973).
  • 该模型实现了0.0027的等错率 (EER),表明了强大的检测能力.

结论:

  • 开发的模型在检测视频流中的暴力行为方面取得了重大进展.
  • 这项研究有助于增强人工智能系统的实时视频监控和预防犯罪.
  • 这些发现对通过智能监控解决方案改善公共安全具有实际意义.