A Bearing Fault Diagnosis Framework Based on Few-Shot Learning with Distribution Consistency and Structural Reparameterization

This is our implemented source code for the paper "A Bearing Fault Diagnosis Framework Based on Few-Shot Learning with Distribution Consistency and Structural Reparameterization" accepted by International Conference on Green Technology and Sustainable Development 7th

Methodology

Environment

conda create -n BEARING python=3.10.12 -y
conda activate BEARING
pip install --upgrade pip
pip install -r requirements.txt

Dataset

CWRU Download Link

Getting Started

Installation

git clone https://github.com/giabao804/few-shot-structural-rep.git
cd few-shot-structural-rep

Training

chmod +x train.sh

• 1-shot training

bash train.sh 1 

• 5-shot training

bash train.sh 5

Testing

chmod +x test.sh

• 1-shot testing

bash test.sh 1 

• 5-shot testing

bash test.sh 5

Contact

Please feel free to contact me via email bao.tg212698@sis.hust.edu.vn or giabaotruong.work@gmail.com if you need anything related to this repo!

Citation

If you feel this code is useful, please give us 1 ⭐ and cite our paper.

@inproceedings{truong2024bearing,
  title={A Bearing Fault Diagnosis Framework Based on Few-Shot Learning with Distribution Consistency and Structural Reparameterization},
  author={Truong, Gia-Bao and Than, Nhu-Linh and Vu, Manh-Hung and Pham, Van-Truong and Nguyen, Thi-Hue and Tran, Thi-Thao},
  booktitle={International Conference on Green Technology and Sustainable Development},
  pages={162--172},
  year={2024},
  organization={Springer}
}