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

Rapid Spur Gear Profile Inspection Using Chromatic Confocal Sensors.

Sensors (Basel, Switzerland)·2026
Same author

Improving Efficacy and Reducing Systemic Toxicity: An In Vitro Study on the Role of Electrospun Gelatin Nanofiber Membrane for Localized Melanoma Treatment.

Bioengineering (Basel, Switzerland)·2025
Same author

An ODE based neural network approach for PM2.5 forecasting.

Scientific reports·2025
Same author

High-quality chromosome-level genome of three Meretrix species using Nanopore and Hi-C technologies.

Scientific data·2025
Same author

Improving AI object detection in fire scenes through data augmentation.

Journal of occupational and environmental hygiene·2025
Same author

Enhancing PM2.5 prediction by mitigating annual data drift using wrapped loss and neural networks.

PloS one·2025

相关实验视频

Updated: Jul 11, 2025

Long-term Behavioral Tracking of Freely Swimming Weakly Electric Fish
10:56

Long-term Behavioral Tracking of Freely Swimming Weakly Electric Fish

Published on: March 6, 2014

12.6K

使用lite 3D全卷积网络识别水下生物的异常行为.

Jung-Hua Wang1,2, Te-Hua Hsu3,4, Yi-Chung Lai5,6

  • 1Deptartment of Electrical Engineering, National Taiwan Ocean University, Keelung City, 20224, Taiwan. jhwang@email.ntou.edu.tw.

Scientific reports
|November 17, 2023
PubMed
概括

这项研究介绍了Lite3D,这是一种轻量级的深度学习模型,用于实时识别异常的水下生物行为. 它通过在水下车辆上的边缘计算提供高效的海洋生态系统健康监测.

更多相关视频

Long-term Video Tracking of Cohoused Aquatic Animals: A Case Study of the Daily Locomotor Activity of the Norway Lobster Nephrops norvegicus
05:57

Long-term Video Tracking of Cohoused Aquatic Animals: A Case Study of the Daily Locomotor Activity of the Norway Lobster Nephrops norvegicus

Published on: April 8, 2019

6.9K
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

501

相关实验视频

Last Updated: Jul 11, 2025

Long-term Behavioral Tracking of Freely Swimming Weakly Electric Fish
10:56

Long-term Behavioral Tracking of Freely Swimming Weakly Electric Fish

Published on: March 6, 2014

12.6K
Long-term Video Tracking of Cohoused Aquatic Animals: A Case Study of the Daily Locomotor Activity of the Norway Lobster Nephrops norvegicus
05:57

Long-term Video Tracking of Cohoused Aquatic Animals: A Case Study of the Daily Locomotor Activity of the Norway Lobster Nephrops norvegicus

Published on: April 8, 2019

6.9K
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

501

科学领域:

  • 海洋生物学 海洋生物学
  • 计算机科学 计算机科学
  • 人工智能的人工智能

背景情况:

  • 海洋健康受到全球变暖和污染的威胁,影响海洋息地和物种.
  • 海洋生物的异常行为是评估海洋健康的关键指标.
  • 目前用于行为识别的深度学习模型是准确的,但计算密集且缓慢.

研究的目的:

  • 为水下生物开发一个实时异常行为识别系统.
  • 创建一个轻量级的深度学习模型,适用于海上车辆边缘计算.
  • 提高用于生态监测的海洋动物行为分析的效率和速度.

主要方法:

  • 整合一个轻量级的深度学习模型 (Lite3D) 与对象检测和多目标跟踪.
  • Lite3D利用感兴趣的区域 (ROI) 和3D卷积来高效地提取特征.
  • 该模型避免了完全连接的层,减少了计算复杂性和大小.

主要成果:

  • 与其他3D模型相比,Lite3D的可训练参数小50倍,轻57倍.
  • 在异常行为识别方面获得了99%的F1分数.
  • 已证明适用于远程操作车辆 (ROV) 或自动水下车辆 (AUV) 的实时边缘计算.

结论:

  • Lite3D提供了一种高效准确的解决方案,用于在海洋环境中实时识别异常行为.
  • 该模型的轻量级设计使其能够部署在水下车辆上进行现场数据处理.
  • 这种方法促进了利用海洋动物行为作为生物计的海洋健康的持续监测.