MED-NAS-Bench: Towards a Multi-task Neural Architecture Search Benchmark for Advancing Medical Imaging Analysis
Hadjer Benmeziane · Kaoutar El Maghraoui · Hamza Ouarnoughi · Smail Niar
Abstract
We introduce MED-NAS-Bench, a comprehensive Neural Architecture Search (NAS) benchmark for medical imaging analysis. This benchmark encompasses task-specific metrics and hardware efficiency assessment, addressing the field's lack of standardized evaluation platforms. It empowers informed decision-making and optimized AI model adoption in real-world healthcare applications leveraging electronic health records (EHR).
Chat is not available.
Successful Page Load