Table 2.
The details of recovery CPU Time of practical wearable ECG datasets (CR = 60%). These results show that the proposed BSBL-ADMM algorithm is robust for the recovery speed for different ECG recordings.
| Rec1 | Rec2 | Rec3 | Rec4 | Rec5 | Rec6 | Rec7 | Rec8 | Mean (±Std) | |
|---|---|---|---|---|---|---|---|---|---|
| BSBL-ADMM | 0.0639 | 0.0635 | 0.0631 | 0.0625 | 0.0625 | 0.0624 | 0.0624 | 0.0628 | 0.0629 (±0.0006) |
| BSBL-BO [25] | 0.2359 | 0.2361 | 0.2353 | 0.2352 | 0.2355 | 0.2350 | 0.2350 | 0.2358 | 0.2355 (±0.0004) |
| BSBL-L1 [25] | 0.0294 | 0.0293 | 0.0291 | 0.0290 | 0.0291 | 0.0291 | 0.0290 | 0.0291 | 0.0292 (±0.0001) |
| BSBL-FM [27] | 0.1605 | 0.2085 | 0.2868 | 0.2541 | 0.2170 | 0.2038 | 0.2826 | 0.2590 | 0.2340 (±0.0044) |
| AIHT [19] | 0.0458 | 0.0377 | 0.0372 | 0.0444 | 0.0425 | 0.0459 | 0.0407 | 0.0414 | 0.0419 (±0.0034) |