Repository containing the Unsupervised Domain Adaptation project developed for the Deep Learning course of the master's degree in Computer Science at University of Trento
Code and experiments for my submission to the BirdCLEF+ 2025 Kaggle competition.
This study audits and mitigates fairness issues in cardiac MRI segmentation across SIEMENS, Philips, and GE scanners. A baseline 2D U-Net showed spurious vendor bias, particularly for the minority GE domain. Implementing a Domain Adversarial Neural Network reduced F1-Score disparity, stabilizing recall and improving clinical safety.
Domain Adaptation With Domain-Adversarial Training of Neural Networks
Domain adversarial network trained on MNIST-M, SVHN, and USPS
Add a description, image, and links to the domain-adversarial-neural-networks topic page so that developers can more easily learn about it.
To associate your repository with the domain-adversarial-neural-networks topic, visit your repo's landing page and select "manage topics."