ECG biometric recognition on ECG-ID
99.56AccuracyAttention-Guided Fusion of 1D and 2D CNNs
Evaluation results
| Method | Links | |
|---|---|---|
| Attention-Guided Fusion of 1D and 2D CNNsOptimal α=0.623, Architecture=InceptionTime + ResNet-34, Segmentation=P–T wave2026.05 | 99.56 | |
| CNN + Self-Supervised Contrastive LearningClass=1D, Year=20242026.05 | 98.77 | |
| ECGXtractorClass=1D, Year=20232026.05 | 97.75 | |
| One-Dimensional Shallow Neural NetworkClass=1D, Year=20232026.05 | 94.39 | |
| LSTMClass=1D, Year=20212026.05 | 93.11 | |
| Siamese Neural NetworksClass=2D, Year=20212026.05 | 86.47 |