CAD classification on Noisy PCG dataset Subject Level 1.0
78.4AccuracyMFCC Conformer
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| MFCC ConformerInput=Denoised, Noise segment rejection=true, Learning=Hybrid Contrastive, Decision Rule=Majority Vote2026.01 | 78.4 | 78.2 | 81.9 | 74.5 | 79.9 | 76.4 | 0.57 | |
| Wav2Vec 2.0Input=Noisy, Decision Rule=Majority Vote2026.01 | 77.1 | 74.3 | 86.5 | 62 | 82.3 | 67.1 | 0.51 | |
| MFCC ConformerInput=Noisy, Noise segment rejection=false, Decision Rule=Majority Vote2026.01 | 74.3 | 73.9 | 80.9 | 66.9 | 76.8 | 70.6 | 0.49 |