ECG Biometrics with ArcFace-Inception: External Validation on MIMIC and HEEDB
Scagnetto A · 2026
Read ECG Biometrics with ArcFace-Inception: External Validation on MIMIC and HEEDB by Scagnetto A (2026) 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-2604-04485.txt.
Description
Scagnetto evaluates an ArcFace-trained 1D Inception-v1 model for ECG biometrics under large galleries, domain shift and multi-year gaps. The model trains on 164,440 12-lead ECGs from 53,079 patients. It is tested on MIMIC-IV-ECG and HEEDB cohorts. Under general comparability it reaches Rank@1 near 0.95 on ASUGI-DB, lower on MIMIC-GC and HEEDB-GC. In a temporal stress test at constant gallery size, Rank@1 falls from 1 to 5 years. ECG identity stays measurable at large scale, but quality hinges on domain shift, drift, gallery size and score processing.
Catalog record
| CREATOR | Scagnetto A |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 42666 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.04485