Motherboard Production Defect Dataset

This repo contains a dataset of motherboard production defects for object detection models, this is currently the biggest public dataset on motherboard defects (as of Dec. 2022). This dataset is built for YOLOv5 (PyTorch), YOLOv7 (PyTorch), YOLOv8 and OpenAI CLIP.

Download links

YOLOv5-PyTorch (221.78MB)
YOLOv7-PyTorch (221.78MB)
YOLOv8 (222.62MB)
CLIP (232.3MB)

Sample Images

Some sample images from the dataset are available in the sample_images folder.

Dataset Stats

Num. of Images	Num. of Annotations	Num. of Classes	Avg. Image Size	Median Image Ratio
389	2860	11	1.10MP	1044x1074

Class Distribution

Citation

If you find this dataset useful in your project, please consider citing our work:

 @misc{yuelin xin_zihan zhou_yuxuan xia_jiuqiang li_qiyu ou_2023,
	title={Computer Motherboard Production Defects},
	url={https://www.kaggle.com/ds/2764127},
	DOI={10.34740/KAGGLE/DS/2764127},
	publisher={Kaggle},
	author={Yuelin Xin and Zihan Zhou and Yuxuan Xia and Jiuqiang Li and Qiyu Ou},
	year={2023}
}