2nd FORTHEM Annual Conference - Connecting Worlds of Science and Society"

Usage of Neural Networks U-Net & GANs in image segmentation, a study case with application

31 Jan 2025, 11:30

15m

ONLINE

Oral presentation Natural and Life Sciences: insights from next-gen researchers Natural and Life Sciences: insights from next-gen researchers

Speaker

Georgian-Cristian Chivu (Universitatea "Lucian Blaga" din Sibiu)

Description

This study examines U-Net and GAN architectures for image segmentation, focusing on remote sensing applications. U-Net excelled in structural segmentation, while GANs proved superior in texture reconstruction. Their integration offers a powerful solution for precise and complex segmentation tasks.