Academic project
Fabric Defect Segmentation Using Convolutional Neural Networks
Study of U-Net semantic segmentation for textile defects using the AITEX dataset and the Dice metric.
Automated fabric inspection must handle a wide variety of textile patterns and defect types. The project investigated convolutional neural networks for semantic segmentation of defective regions in fabric images.
The Lattes description identifies U-Net, the public AITEX image dataset, and the Dice metric as components of the proposed experiments. The record marks the project as completed in 2022, but does not provide quantitative findings that could be attributed to this work.
Team listed in Lattes: Jefferson Oliveira Andrade, Karin Satie Komati (project lead), Kelly Assis de Souza Gazolli, Daniel Ribeiro Trindade. The Lattes record reports that the project was completed in 2022. Support listed in Lattes: Fundação de Amparo à Pesquisa do Espírito Santo (scholarship).