Search Results for author:
Found 3 papers, 2 papers with code
Rethinking Model Prototyping through the MedMNIST+ Dataset Collection
The integration of deep learning based systems in clinical practice is often impeded by challenges rooted in limited and heterogeneous medical datasets.
Integrating kNN with Foundation Models for Adaptable and Privacy-Aware Image Classification
Traditional deep learning models implicity encode knowledge limiting their transparency and ability to adapt to data changes.
unORANIC: Unsupervised Orthogonalization of Anatomy and Image-Characteristic Features
We introduce unORANIC, an unsupervised approach that uses an adapted loss function to drive the orthogonalization of anatomy and image-characteristic features.
Cannot find the paper you are looking for? You can
Submit a new open access paper.
