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

相关实验视频

Updated: Jul 5, 2025

Evaluating the Effect of Roadside Parking on a Dual-Direction Urban Street
14:55

Evaluating the Effect of Roadside Parking on a Dual-Direction Urban Street

Published on: January 20, 2023

3.3K

为使用基础设施传感器启用设备的弱势道路使用者提供了一个增强的探测器.

Jian Shi1, Dongxian Sun1, Minh Kieu2

  • 1School of Mechanical Engineering, Beijing Institute of Technology, Beijing 100081, China.

Sensors (Basel, Switzerland)
|January 11, 2024
PubMed
概括
此摘要是机器生成的。

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

Evaluation of cold tolerance in maize diversity panel and transcriptome analysis of two extreme genotypes under cold stress.

Plant physiology and biochemistry : PPB·2025
Same author

Adaptive occlusion object detection algorithm based on OL-IoU.

Scientific reports·2024
Same author

Multi-modality trajectory prediction with the dynamic spatial interaction among vehicles under connected vehicle environment.

Scientific reports·2024
Same author

VV-YOLO: A Vehicle View Object Detection Model Based on Improved YOLOv4.

Sensors (Basel, Switzerland)·2023
Same author

Transcriptomic and metabolomic profiling provide novel insights into fruit development and flesh coloration in Prunus mira Koehne, a special wild peach species.

BMC plant biology·2019
Same author

Genome-wide Dissection of Co-selected UV-B Responsive Pathways in the UV-B Adaptation of Qingke.

Molecular plant·2019

本研究引入了使用基础设施传感器对脆弱道路使用者 (VRU) 进行改进的探测器. 改进的模型实现了更好的检测精度和更快的实时性能,用于智能交通系统.

科学领域:

  • 计算机视觉 计算机视觉
  • 人工智能的人工智能
  • 智能运输系统 智能运输系统

背景情况:

  • 精确和实时检测易受伤害的道路使用者 (VRU) 对于智能交通监控至关重要.
  • 现有的VRU检测方法面临效率低下的问题,特别是高分辨率图像中的小目标.

研究的目的:

  • 开发一个增强的探测器,以使用基础设施传感器启用设备进行准确和高效的VRU检测.
  • 为了改善小目标的特征提取,并减少实时应用的模型复杂性.

主要方法:

  • 使用轻量级的骨干网络,具有无参数的注意力机制,用于增强功能提取.
  • 实现了精简的"子"和动态检测头,并采用了用于模型压缩的修剪算法.
  • 在使用De_VRU数据集的Hisilicon_Hi3516DV300平台上部署模型.
  • 在BDD100K和LLVIP数据集上使用YOLOv7-tiny作为基线进行了废弃研究.

主要成果:

  • 在mAP@50指标上实现了超过12%的改进.
  • 模型参数数量减少了40%以上.
  • 推断时间减少了50%,使实时检测成为可能.
关键词:
在VRU检测检测.基础设施传感器支持的工程.模型轻量级的轻量级模型对象检测检测对象检测对象检测

更多相关视频

Design and Analysis for Fall Detection System Simplification
08:05

Design and Analysis for Fall Detection System Simplification

Published on: April 6, 2020

10.7K
Evaluation of an Exclusive Spur Dike U-Turn Design with Radar-Collected Data and Simulation
11:41

Evaluation of an Exclusive Spur Dike U-Turn Design with Radar-Collected Data and Simulation

Published on: February 1, 2020

20.4K

相关实验视频

Last Updated: Jul 5, 2025

Evaluating the Effect of Roadside Parking on a Dual-Direction Urban Street
14:55

Evaluating the Effect of Roadside Parking on a Dual-Direction Urban Street

Published on: January 20, 2023

3.3K
Design and Analysis for Fall Detection System Simplification
08:05

Design and Analysis for Fall Detection System Simplification

Published on: April 6, 2020

10.7K
Evaluation of an Exclusive Spur Dike U-Turn Design with Radar-Collected Data and Simulation
11:41

Evaluation of an Exclusive Spur Dike U-Turn Design with Radar-Collected Data and Simulation

Published on: February 1, 2020

20.4K

结论:

  • 改进的探测器在VRU检测的准确性和效率方面取得了显著的改进.
  • 该模型的紧架构和快速推断时间使其适合在智能交通系统中实际部署.
  • 该研究强调了先进的深度学习技术在提高道路安全方面的潜力.