Data

Tools

Indexes

Applications

Experiments

Open Source Science

dcf-yolov8

https://github.com/sprboot/dcf-yolov8

Science Score: 54.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • ✓
    CITATION.cff file
    Found CITATION.cff file
  • ✓
    codemeta.json file
    Found codemeta.json file
  • ✓
    .zenodo.json file
    Found .zenodo.json file
  • ○
    DOI references
  • ✓
    Academic publication links
    Links to: mdpi.com
  • ○
    Academic email domains
  • ○
    Institutional organization owner
  • ○
    JOSS paper metadata
  • ○
    Scientific vocabulary similarity
    Low similarity (7.3%) to scientific vocabulary
Last synced: 10 months ago · JSON representation ·

Repository

Basic Info
  • Host: GitHub
  • Owner: SprBoot
  • Language: Python
  • Default Branch: master
  • Size: 150 MB
Statistics
  • Stars: 2
  • Watchers: 1
  • Forks: 1
  • Open Issues: 1
  • Releases: 0
Created over 2 years ago · Last pushed over 2 years ago
Metadata Files
Readme Contributing Citation Security

docs/README.md

Code description

This is the public code of the paper DCF-Yolov8: An Improved Algorithm for Aggregating Low-Level Features to Detect Agricultural Pests and Diseases The paper address is https://www.mdpi.com/2073-4395/13/8/2012

Paper algorithm improvement

Based on the characteristics of the IP102 agricultural pest data set, we improved the C2F module. The module structure is shown in the figure below. This module has better feature extraction effects on low-level features of agricultural pests and diseases.

image

For specific codes, see DCF-Yolov8/ultralytics/nn/modules/block.py class C2F

We use the Mish activation function, which is more robust in dealing with nonlinear features. Our algorithm's ability to learn feature effects is still strong at higher rounds.

image

For details on modifying the activation function code, see DCF-Yolov8/ultralytics/nn/modules/conv.py

We have uploaded the model training results.You can download it at https://drive.google.com/drive/folders/1q5YWRY2el8SOZOdxBRwH-PA7rsCuoPjm?usp=drive_link. Ours.zip and Ours-val.zip are the training results and verification results of our improved algorithm. yolov8.zip and yolov8-va.zip are the training and verification results of the yolov8 algorithm. yolov8-mish.zip is the result we used for ablation experiments. The results include our model weights and experimental metrics.

We have summarized the original images in the paper, see agronomy.pptx for details

Acknowledgments

Our paper is an improvement on the Yolov8 algorithm, thanks to the Yolov8 team[https://github.com/ultralytics/ultralytics.git]. Thanks to the IP102 dataset team[https://github.com/xpwu95/IP102.git].Thanks to yolox[https://github.com/Megvii-BaseDetection/YOLOX.git], yolov5 algorithm[https://github.com/ultralytics/yolov5.git]. Heat maps are a very effective and intuitive way to demonstrate the effect of our algorithmsthe code for the heat map comes from https://github.com/z1069614715/objectdetection_script.git.

Owner

  • Login: SprBoot
  • Kind: user

Citation (CITATION.cff) 

cff-version: 1.2.0
preferred-citation:
  type: software
  message: If you use this software, please cite it as below.
  authors:
  - family-names: Jocher
    given-names: Glenn
    orcid: "https://orcid.org/0000-0001-5950-6979"
  - family-names: Chaurasia
    given-names: Ayush
    orcid: "https://orcid.org/0000-0002-7603-6750"
  - family-names: Qiu
    given-names: Jing
    orcid: "https://orcid.org/0000-0003-3783-7069"
  title: "YOLO by Ultralytics"
  version: 8.0.0
  # doi: 10.5281/zenodo.3908559  # TODO
  date-released: 2023-1-10
  license: AGPL-3.0
  url: "https://github.com/ultralytics/ultralytics"

GitHub Events

Total
  • Watch event: 1
  • Fork event: 1
Last Year
  • Watch event: 1
  • Fork event: 1