CoChat
Sign Up
Cited by 54Smart Agriculture and AIAdvanced Neural Network ApplicationsIndustrial Vision Systems and Defect Detection

The DOI matches its Crossref record; 38 of the 38 references it cites were found in a registry or an index.

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

Open access
Mwitta, Canicius, Rains, Glen C., Prostko, Eric·2024·Sensors
DOI 10.3390/s24020514PMC PMC10819251
Open this paper in CoChat — free accountOpen PDF

Abstract

The knowledge that precision weed control in agricultural fields can reduce waste and increase productivity has led to research into autonomous machines capable of detecting and removing weeds in real time. One of the driving factors for weed detection is to develop alternatives to herbicides, which are becoming less effective as weed species develop resistance. Advances in deep learning technology have significantly improved the robustness of weed detection tasks. However, deep learning algorithms often require extensive computational resources, typically found in powerful computers that are not suitable for deployment in robotic platforms. Most ground rovers and UAVs utilize embedded computers that are portable but limited in performance. This necessitates research into deep learning models that are computationally lightweight enough to function in embedded computers for real-time applications while still maintaining a base level of detection accuracy. This paper evaluates the weed detection performance of three real-time-capable deep learning models, YOLOv4, EfficientDet, and CenterNet, when run on a deep-learning-enabled embedded computer, an Nvidia Jetson Xavier AGX. We tested the accuracy of the models in detecting 13 different species of weeds and assesses their real-time viability through their inference speeds on an embedded computer compared to a powerful deep learning PC. The results showed that YOLOv4 performed better than the other models, achieving an average inference speed of 80 ms per image and 14 frames per second on a video when run on an imbedded computer, while maintaining a mean average precision of 93.4% at a 50% IoU threshold. Furthermore, recognizing that some real-world applications may require even greater speed, and that the detection program would not be the only task running on the embedded computer, a lightweight version of the YOLOv4 model, YOLOv4-tiny, was tested for improved performance in an embedded computer. YOLOv4-tiny impressively achieved an average inference speed of 24.5 ms per image and 52 frames per second, albeit with a slightly reduced mean average precision of 89% at a 50% IoU threshold, making it an ideal choice for real-time weed detection.

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.3390/s24020514

Cites

38 of 38 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 Examples of images of 13 weed species.
Figure 1 Examples of images of 13 weed species.CC BY · PubMed Central
Figure 2 Example of labeling weed image using LabelImg.
Figure 2 Example of labeling weed image using LabelImg.CC BY · PubMed Central
Figure 3 Labeled dataset split.
Figure 3 Labeled dataset split.CC BY · PubMed Central
Figure 4 YOLOv4 training on darknet platform.
Figure 4 YOLOv4 training on darknet platform.CC BY · PubMed Central
Figure 5 EfficientDet: total loss against number of training steps—training loss (orange) and validation loss (blue).
Figure 5 EfficientDet: total loss against number of training steps—training loss (orange) and validation loss (blue).CC BY · PubMed Central
Figure 6 EfficientDet: precision (mAP@0.5) against number of training steps.
Figure 6 EfficientDet: precision (mAP@0.5) against number of training steps.CC BY · PubMed Central
Figure 7 EfficientDet: recall against number of training steps.
Figure 7 EfficientDet: recall against number of training steps.CC BY · PubMed Central
Figure 8 CenterNet: total loss against number of training steps—training loss (orange) and validation loss (blue).
Figure 8 CenterNet: total loss against number of training steps—training loss (orange) and validation loss (blue).CC BY · PubMed Central
Figure 9 CenterNet: precision (mAP@0.5) against number of training steps.
Figure 9 CenterNet: precision (mAP@0.5) against number of training steps.CC BY · PubMed Central
Figure 10 CenterNet: recall against number of training steps.
Figure 10 CenterNet: recall against number of training steps.CC BY · PubMed Central
Figure 11 Nvidia Jetson Xavier AGX.
Figure 11 Nvidia Jetson Xavier AGX.CC BY · PubMed Central
Figure 12 Precision, recall, and IoU illustration.
Figure 12 Precision, recall, and IoU illustration.CC BY · PubMed Central
Figure 13 YOLOv4 detection—detected 3 weeds.
Figure 13 YOLOv4 detection—detected 3 weeds.CC BY · PubMed Central
Figure 14 CenterNet detection—detected 3 weeds.
Figure 14 CenterNet detection—detected 3 weeds.CC BY · PubMed Central
Figure 15 EfficientDet detection—detected all 4 weeds.
Figure 15 EfficientDet detection—detected all 4 weeds.CC BY · PubMed Central
Figure 16 Single purple nutsedge plant detected by CenterNet.
Figure 16 Single purple nutsedge plant detected by CenterNet.CC BY · PubMed Central
Figure 17 Single purple nutsedge plant detected by EfficientDet.
Figure 17 Single purple nutsedge plant detected by EfficientDet.CC BY · PubMed Central
Figure 18 Two purple nutsedge plants detected by YOLOv4.
Figure 18 Two purple nutsedge plants detected by YOLOv4.CC BY · PubMed Central
Figure 19 YOLOv4 confusion matrix on the test dataset.
Figure 19 YOLOv4 confusion matrix on the test dataset.CC BY · PubMed Central
Figure 20 EfficientDet confusion matrix on the test dataset.
Figure 20 EfficientDet confusion matrix on the test dataset.CC BY · PubMed Central
Figure 21 CenterNet confusion matrix on the test dataset.
Figure 21 CenterNet confusion matrix on the test dataset.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 38 references

Research works printed but not linked: 3

  1. Rueda-Ayala, V., Rasmussen, J., and Gerhards, R. (2010). Precision Crop Protection—the Challenge and Use of Heterogeneity, Springer.
  2. Zhu, W., and Zhu, X. (2009, January 20–22). The Application of Support Vector Machine in Weed Classification. Proceedings of the 2009 IEEE International Conference on Intelligent Computing and Intelligent Systems, Shanghai, China.
  3. Bochkovskiy, A., Wang, C.-Y., and Liao, H.-Y.M. (2020). YOLOv4: Optimal Speed and Accuracy of Object Detection. arXiv.

Cited by

Sorted by citations

See all 54

Related

See all

Metadata

Authors
Mwitta, C., Rains, G. C., & Prostko, E.
Affiliations
University of Georgia (US)
Corresponding author
  • Mwitta, Canicius
  • Rains, Glen C.
ORCID
Venue
Sensors
Publisher
MDPI AG
Published
2024-01-14
Volume
24
Issue
2
Pages
514
ISSN
1424-8220 (ISSN-L)
DOI
10.3390/s24020514
PMC
PMC10819251
OpenAlex
W4390879300
Cited by
41
References
38
Citations by year
2026: 12 · 2025: 19 · 2024: 10
FWCI
10.01
Type
journal article
Language
en
Open access
Licence
Locations
  • Sensors (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
  • US Cotton Incorporated
  • US Georgia Peanut Commission
Topics
  • Life Sciences › Agricultural and Biological Sciences › Plant Science › Smart Agriculture and AI
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
  • Physical Sciences › Engineering › Industrial and Manufacturing Engineering › Industrial Vision Systems and Defect Detection
Keywords
weed detection, real-time object detection, embedded computing, YOLOv4, YOLOv4-tiny, EfficientDet, CenterNet, lightweight deep learning, autonomous agricultural robots, precision weed control, inference speed, mean average precision

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Evaluation of Inference Performance of Deep Learning Models for Real-Time Weed Detection in an Embedded Computer. https://app.cochat.ai/papers/10.3390/s24020514
The paper
Mwitta, C., Rains, G. C., & Prostko, E. (2024). Evaluation of Inference Performance of Deep Learning Models for Real-Time Weed Detection in an Embedded Computer. Sensors. https://doi.org/10.3390/s24020514