Skip to content

Efficient Local Attention Modeling for High-Performance Real-Time Insulator Defect Detection

License

Notifications You must be signed in to change notification settings

joshua-atolagbe/YOLO-ELA

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

38 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

YOLO-ELA: Efficient Local Attention Modeling for High-Performance Real-Time Insulator Defect Detection

Architecture

Dataset

  • Training and validation data can be downloaded here
  • Test data can be downloaded here.

Requirements

Install requirements

pip install -r requirements.txt
  • +NVIDIA GPU + CUDA CuDNN
  • +Linux (Ubuntu)
  • +Python 3.10

Get the test results

  • Download YOLO-ELA checkpoint here
  • Open a terminal and run
python test.py\
        --model 'models/yolo_ela.pt'\ 
       

This automatically create a new directory called run. Navigate to see results

How to train YOLO-ELA

  • Open terminal and run
python train.py \
        --model_scale 'ela' \
        --cfg 'ultralytics/cfg/data.yaml' \
        --aug True \
        --name 'ela' \
        --epochs 100 \
        --bs 16 \
        --img_sz 640 # Image size can either be 320 or 640

Citing

@article{yoloela,
  author = {Olalekan Akindele and Joshua Atolagbe},
  title = {YOLO-ELA: Efficient Local Attention Modeling for High-Performance Real-Time Insulator Defect Detection},
  journal = {arXiv preprint arXiv:2410.11727},
  year = {2024}
}

Credit

The codes in this repository are based on Ultralytics

About

Efficient Local Attention Modeling for High-Performance Real-Time Insulator Defect Detection

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published