Investigating Shift Equivalence of Convolutional Neural Networks in Industrial Defect Segmentation

1 code implementation • 29 Sep 2023 • Zhen Qu, Xian Tao, Fei Shen, Zhengtao Zhang, Tao Li

In industrial defect segmentation tasks, while pixel accuracy and Intersection over Union (IoU) are commonly employed metrics to assess segmentation performance, the output consistency (also referred to equivalence) of the model is often overlooked.

Data Augmentation Segmentation

The Second-place Solution for CVPR VISION 23 Challenge Track 1 -- Data Effificient Defect Detection

1 code implementation • 25 Jun 2023 • Xian Tao, Zhen Qu, Hengliang Luo, Jianwen Han, Yonghao He, Danfeng Liu, Chengkan Lv, Fei Shen, Zhengtao Zhang

The Vision Challenge Track 1 for Data-Effificient Defect Detection requires competitors to instance segment 14 industrial inspection datasets in a data-defificient setting.

Defect Detection Instance Segmentation +2

Deep Learning for Unsupervised Anomaly Localization in Industrial Images: A Survey

no code implementations • 21 Jul 2022 • Xian Tao, Xinyi Gong, Xin Zhang, Shaohua Yan, Chandranath Adak

This paper aims to help researchers in this field by comprehensively surveying recent achievements in unsupervised anomaly localization in industrial images using deep learning.