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

Effects of aluminate flocculant on turion germination and seedling growth of Potamogeton crispus.

Aquatic toxicology (Amsterdam, Netherlands)·2017
Same author

Folate-conjugated amphiphilic block copolymer micelle for targeted and redox-responsive delivery of doxorubicin.

Journal of biomaterials science. Polymer edition·2017
Same author

Survival of the fittest: Cancer challenges T cell metabolism.

Cancer letters·2017
Same author

EEG-based emotion estimation using adaptive tracking of discriminative frequency components.

Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual International Conference·2017
Same author

Two new stilbenoids from aerial parts of Flickingeria fimbriata.

Journal of Asian natural products research·2017
Same author

Enhancing performance of P300-Speller under mental workload by incorporating dual-task data during classifier training.

Computer methods and programs in biomedicine·2017

相关实验视频

Updated: Jun 15, 2025

Study on the Metabolism of Six Systemic Insecticides in a Newly Established Cell Suspension Culture Derived from Tea Camellia Sinensis L. Leaves
10:35

Study on the Metabolism of Six Systemic Insecticides in a Newly Established Cell Suspension Culture Derived from Tea Camellia Sinensis L. Leaves

Published on: June 15, 2019

7.7K

TP-Transfiner:茶叶害虫的高质量细分网络.

Ruizhao Wu1, Feng He1,2, Ziyang Rong1,2

  • 1College of Informatics, Huazhong Agricultural University, Wuhan, China.

Frontiers in plant science
|August 28, 2024
PubMed
概括

本研究介绍了TeaPest-Transfiner (TP-Transfiner),这是一个先进的AI模型,用于检测和细分茶叶害虫,提高挑战密集和模仿场景的准确性. 新的框架增强了特征提取,并实现了最先进的性能,有助于有效地控制茶叶生产的质量.

关键词:
面具转精器 面具转精器注意力机制注意力机制密集和模仿场景的情况.实例细分 实例细分 实例细分茶叶虫害害虫害 在茶叶虫害害害虫害害虫害虫害虫害虫害害虫害虫害

更多相关视频

Remote Sensing Evaluation of Two-spotted Spider Mite Damage on Greenhouse Cotton
05:03

Remote Sensing Evaluation of Two-spotted Spider Mite Damage on Greenhouse Cotton

Published on: April 28, 2017

8.5K
Isolation and Selection of Entomopathogenic Fungi from Soil Samples and Evaluation of Fungal Virulence against Insect Pests
09:42

Isolation and Selection of Entomopathogenic Fungi from Soil Samples and Evaluation of Fungal Virulence against Insect Pests

Published on: September 28, 2021

9.2K

相关实验视频

Last Updated: Jun 15, 2025

Study on the Metabolism of Six Systemic Insecticides in a Newly Established Cell Suspension Culture Derived from Tea Camellia Sinensis L. Leaves
10:35

Study on the Metabolism of Six Systemic Insecticides in a Newly Established Cell Suspension Culture Derived from Tea Camellia Sinensis L. Leaves

Published on: June 15, 2019

7.7K
Remote Sensing Evaluation of Two-spotted Spider Mite Damage on Greenhouse Cotton
05:03

Remote Sensing Evaluation of Two-spotted Spider Mite Damage on Greenhouse Cotton

Published on: April 28, 2017

8.5K
Isolation and Selection of Entomopathogenic Fungi from Soil Samples and Evaluation of Fungal Virulence against Insect Pests
09:42

Isolation and Selection of Entomopathogenic Fungi from Soil Samples and Evaluation of Fungal Virulence against Insect Pests

Published on: September 28, 2021

9.2K

科学领域:

  • 农业科学 农业科学
  • 计算机视觉 计算机视觉
  • 人工智能的人工智能

背景情况:

  • 传统的卷积神经网络 (CNN) 方法难以准确有效地检测茶叶害虫,特别是在密集或模拟场景中,因为特征提取不足.
  • 迅速检测和控制茶叶害虫对于保持茶叶生产质量至关重要.

研究的目的:

  • 开发一个端到端的框架,TeaPest-Transfiner (TP-Transfiner),用于改进茶叶害虫检测和细分,特别是解决密集和模仿环境中的挑战.
  • 增强功能提取能力,超越传统的CNN模块,以提高害虫识别的准确性和效率.

主要方法:

  • 整合一个可变形的注意力块,结合可变形卷积和自我注意力,以改善特征提取.
  • 增强特征金字塔网络 (FPN) 架构,使用特征对齐金字塔网络 (FaPN).
  • 在训练和数据集特定参数调整期间,利用焦点损失进行样本平衡,同时创建了TeaPestDataset,包含29种茶叶害虫的1752张图像.

主要成果:

  • 在TeaPest数据集上,TP-Transfiner模型实现了最先进的性能,检测精度 (AP50) 为87.211%,分段性能为87.381%.
  • 与Mask R-CNN.相比,在细分平均精度 (mAP) 中显著提高了9.4%.
  • 与Mask R-CNN相比,模型大小减少了30%,同时保持了快速推断速度和紧的模型,表明了实际适用性.

结论:

  • 提议的TP-Transfiner框架有效地解决了传统方法在复杂场景中检测和细分茶叶害虫的局限性.
  • 该模型的增强特征提取,优化架构和数据集特定训练有助于其卓越的性能和实用潜力,用于现实世界茶园害虫控制.