HyDA: Hypernetworks for Test Time Domain Adaptation in Medical Imaging Analysis
Doron Serebro, Tammy Riklin-Raviv · 2025
Read HyDA: Hypernetworks for Test Time Domain Adaptation in Medical Imaging Analysis by Doron Serebro, Tammy Riklin-Raviv (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-04979.txt.
Description
Doron Serebro and Tammy Riklin-Raviv tackle a common failure of medical image models. Their paper came out in two thousand twenty-five. Scans differ between hospitals because machines, protocols, and patients differ. A model trained at one site often drops in accuracy at another. Most fixes need many sample scans from the new site during training. In real clinics, new-site scans may arrive only at test time.
Their framework, HyDA, uses a hypernetwork that adapts the model on the spot. Instead of erasing site traits, it reads them and adjusts to them. Adaptation happens at test time, with no retraining round. The design aims at tools that can travel between hospitals. That matters for care settings where data cannot be gathered in advance.
Catalog record
| CREATOR | Doron Serebro, Tammy Riklin-Raviv |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 19 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-04979