Table 1.
Data distribution of the FluSense data.
| Original | Pre-Processing | |||
|---|---|---|---|---|
| # | Train | Val | ∑ | |
| Breathe | 167 | 238 | 58 | 297 |
| Cough | 2,486 | 6,148 | 1,537 | 7,685 |
| Gasp | 337 | 315 | 79 | 394 |
| Other | 3,863 | 15,059 | 3,765 | 18,824 |
| Silence | 832 | 1,116 | 279 | 1,395 |
| Sneeze | 611 | 540 | 135 | 675 |
| Sniffle | 589 | 604 | 151 | 755 |
| Speech | 2,615 | 16,614 | 4,154 | 20,768 |
| Throat clearing | 102 | 118 | 29 | 147 |
| ∑ | 11,602 | 40,752 | 10,188 | 50,940 |
The “original” column indicates the number of audio samples; whereas the “pre-processing” columns show the number of segments with unified length of 1 s.