ECG-biometrics-bench standardizes evaluation to expose the Random Split Fallacy, where intra-session splits inflate ECG biometric performance, revealing temporal drift degradation that is not model-specific and can be partially mitigated by multi-session template fusion.
Novel fiducial and non-fiducial approaches to electrocardiogram-based biometric systems,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
ECG-biometrics-bench: A Unified Framework for Reproducible Benchmarking of ECG Biometrics
ECG-biometrics-bench standardizes evaluation to expose the Random Split Fallacy, where intra-session splits inflate ECG biometric performance, revealing temporal drift degradation that is not model-specific and can be partially mitigated by multi-session template fusion.