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

Light Acquisition02:16

Light Acquisition

8.6K
In order to produce glucose, plants need to capture sufficient light energy. Many modern plants have evolved leaves specialized for light acquisition. Leaves can be only millimeters in width or tens of meters wide, depending on the environment. Due to competition for sunlight, evolution has driven the evolution of increasingly larger leaves and taller plants, to avoid shading by their neighbors with contaminant elaboration of root architecture and mechanisms to transport water and nutrients.
8.6K

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

Vaccine Ligand Binding Assay Life Cycle Management, Assay Maintenance, Monitoring and Transfer.

Bioanalysis·2026
Same author

Immune correlates analysis in NextCOVE trial for a next-generation mRNA-1283 COVID-19 vaccine.

Human vaccines & immunotherapeutics·2026
Same author

Two decades of research on drug resistance in renal cell carcinoma.

Discover oncology·2026
Same author

Sonophysical mechanisms of bubble fragmentation near an ultrasonically vibrating boundary: influence of frequency and amplitude on interfacial intensification.

Ultrasonics sonochemistry·2026
Same author

Aging and prostate health: meta-analytic insights into age-related prostatic disorders.

Frontiers in oncology·2026
Same author

FAM-related prognostic molecular subtype screening identified epithelial-derived <i>MAOA</i>-inhibiting bladder cancer.

Frontiers in cell and developmental biology·2026

相关实验视频

Updated: Sep 9, 2025

Harmonic Radar Tags for Insect Tracking: Lightweight, Low-cost, and Accessible
14:44

Harmonic Radar Tags for Insect Tracking: Lightweight, Low-cost, and Accessible

Published on: May 13, 2025

1.1K

PKD-YOLOv8:用于轻量级大麻虫检测的合作修剪和知识蒸框架

Haifeng Yu1, Qingting Luo1, Wei Peng1

  • 1College of Information and Intelligence, Hunan Agricultural University, Changsha 410128, China.

Sensors (Basel, Switzerland)
|August 28, 2025
PubMed
概括

这项研究引入了一种轻量级的人工智能模型来检测菜虫害,显著减少了尺寸和计算需求,同时保持了高精度. 优化模型在边缘设备上显示出优异的实时性能, 有助于作物保护.

关键词:
没有.知识的蒸模型修剪油菜的害虫检测

更多相关视频

A Precise and Autonomous System for the Detection of Insect Emergence Patterns
06:22

A Precise and Autonomous System for the Detection of Insect Emergence Patterns

Published on: January 9, 2019

5.8K
Using Insect Electroantennogram Sensors on Autonomous Robots for Olfactory Searches
07:23

Using Insect Electroantennogram Sensors on Autonomous Robots for Olfactory Searches

Published on: August 4, 2014

23.2K

相关实验视频

Last Updated: Sep 9, 2025

Harmonic Radar Tags for Insect Tracking: Lightweight, Low-cost, and Accessible
14:44

Harmonic Radar Tags for Insect Tracking: Lightweight, Low-cost, and Accessible

Published on: May 13, 2025

1.1K
A Precise and Autonomous System for the Detection of Insect Emergence Patterns
06:22

A Precise and Autonomous System for the Detection of Insect Emergence Patterns

Published on: January 9, 2019

5.8K
Using Insect Electroantennogram Sensors on Autonomous Robots for Olfactory Searches
07:23

Using Insect Electroantennogram Sensors on Autonomous Robots for Olfactory Searches

Published on: August 4, 2014

23.2K

科学领域:

  • 农业科学
  • 计算机视觉
  • 人工智能

背景情况:

  • 是一种重要的全球作物,
  • 有效的害虫防治对于保持大麻种子的生产和经济价值至关重要.
  • 现有的检测方法可能缺乏实地应用的效率或实时功能.

研究的目的:

  • 开发一种轻量级,高效的深度学习模型来检测菜虫害.
  • 压缩一个YOLOv8s模型以提高边缘设备的性能.
  • 通过一种新的蒸策略来增强模型的特征表示能力.

主要方法:

  • 使用基于模型分析和灵敏度评估的结构化修剪.
  • 提出了一种新的轻量级蒸模式 (LMGD) 战略,将Logit和MGD蒸整合起来.
  • 为实验验证开发了一套专门的菜虫害数据集 (ACEFP).

主要成果:

  • 在ACEFP数据集中实现了96.7%的mAP@0.5,93.2%的准确性和92.7%的回忆.
  • 压缩模型大小从11.2MB降至4.4MB (减少60.7%),FLOP从28.3G降至10.01G (减少64.6%).
  • 保持高精度,只有0.1%的降低,并实现了11.76 FPS的Jetson纳米实时推断.

结论:

  • 拟议的轻量级协作压缩学习方法有效地减少了模型大小和计算负载.
  • 在没有显著的精度损失的情况下,LMGD蒸策略提高了学生模型的性能.
  • 优化模型显示出优秀的实时病虫检测能力, 适合农业边缘部署.