A Multi-task and Multi-scale Convolutional Neural Networks for Automatic Recognition of Woven Fabric Pattern

Abstract

The figure shows the methodology of the project whose framework has been initially established. The project aims to apply in real production which includes on-loom system, online web service and tracking single yarns based on CNN to detect fabric defects.

Shuo Meng

PhD candidate of Biomedical Engineering.

My research interests include Medical Signal Processing, Deep Learning, Computer Vision and Full-stack Development.