Search research articles

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Development of a Low-Cost Distributed Computing Pipeline for High-Throughput Cotton Phenotyping.

Sensors (Basel, Switzerland)·2024

Same author

A Review of Successes and Impeding Challenges of IoT-Based Insect Pest Detection Systems for Estimating Agroecosystem Health and Productivity of Cotton.

Sensors (Basel, Switzerland)·2023

See all related articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Video

Updated: Jul 5, 2025

Integration of Animal Behavioral Assessment and Convolutional Neural Network to Study Wasabi-Alcohol Taste-Smell Interaction

Integration of Animal Behavioral Assessment and Convolutional Neural Network to Study Wasabi-Alcohol Taste-Smell Interaction

Published on: August 16, 2024

Evaluation of Inference Performance of Deep Learning Models for Real-Time Weed Detection in an Embedded Computer.

Canicius Mwitta^1,2, Glen C Rains², Eric Prostko³

¹College of Engineering, University of Georgia, Athens, GA 30602, USA.

Sensors (Basel, Switzerland)

|January 23, 2024

Summary

This summary is machine-generated.

Researchers evaluated deep learning models for real-time weed detection on embedded computers. YOLOv4 and YOLOv4-tiny demonstrated effective weed identification and processing speeds suitable for agricultural robots.

Keywords:

deep learning weed detection precision weeding weed detection weed detection inference in embedded computer

More Related Videos

Author Spotlight: UAV Remote Sensing for Efficient Invasive Plant Biomass Estimation

Author Spotlight: UAV Remote Sensing for Efficient Invasive Plant Biomass Estimation

Published on: February 9, 2024

Deep Neural Networks for Image-Based Dietary Assessment

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

Related Experiment Videos

Last Updated: Jul 5, 2025

Integration of Animal Behavioral Assessment and Convolutional Neural Network to Study Wasabi-Alcohol Taste-Smell Interaction

Integration of Animal Behavioral Assessment and Convolutional Neural Network to Study Wasabi-Alcohol Taste-Smell Interaction

Published on: August 16, 2024

Author Spotlight: UAV Remote Sensing for Efficient Invasive Plant Biomass Estimation

Author Spotlight: UAV Remote Sensing for Efficient Invasive Plant Biomass Estimation

Published on: February 9, 2024

Deep Neural Networks for Image-Based Dietary Assessment

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

Area of Science:

Agricultural Robotics
Computer Vision
Machine Learning

Background:

Precision weed control is crucial for reducing agricultural waste and increasing productivity.
Developing alternatives to herbicides is necessary due to increasing weed resistance.
Deep learning models offer robust weed detection but often require significant computational resources, limiting their use in portable robotic platforms.

Purpose of the Study:

To evaluate the performance of real-time deep learning models for weed detection on embedded computers.
To compare the accuracy and inference speed of YOLOv4, EfficientDet, and CenterNet on an Nvidia Jetson Xavier AGX.
To assess the viability of lightweight models like YOLOv4-tiny for real-time applications on resource-constrained robotic systems.

Main Methods:

Tested YOLOv4, EfficientDet, and CenterNet on an Nvidia Jetson Xavier AGX embedded computer.
Evaluated detection accuracy for 13 weed species.
Measured inference speeds on the embedded computer and compared them to a powerful deep learning PC.
Included YOLOv4-tiny to assess performance trade-offs between speed and accuracy.

Main Results:

YOLOv4 achieved an average inference speed of 80 ms/image (14 fps) with 93.4% mean average precision (mAP) at 50% IoU on the embedded computer.
YOLOv4-tiny demonstrated significantly faster inference speeds of 24.5 ms/image (52 fps) with a slightly reduced mAP of 89% at 50% IoU.
Both models showed viable real-time performance on the embedded system.

Conclusions:

YOLOv4 offers a strong balance of accuracy and speed for real-time weed detection on embedded systems.
YOLOv4-tiny provides an excellent option when higher processing speeds are critical, with a minimal accuracy compromise.
These findings support the deployment of deep learning-based autonomous weed detection systems in agriculture.