About this document
Adversarial Semi-Supervised Crack Detection by Parisa is a document available to read on EtoBox.
This document presents a semi-supervised method for pavement crack detection using adversarial learning. The method utilizes unlabeled images during training to improve crack detection accuracy compared to methods that only use labeled images. A full convolution discriminator is adopted that can distinguish ground truth segmentations from predictions. The adversarial loss is combined with cross-entropy loss to train the segmentation network. Experimental results on benchmark datasets show the proposed metho
- Author
- Parisa
- Language
- EN