Table 3 The average accuracy and processing time per frames of three algorithms in different scenarios
From: Voice activity detection in the presence of transient based on graph
Algorithm | Â | Two categories | Three categories | Processing time per frame |
|---|---|---|---|---|
PND | Speech | 83.77% | 82.42% | 0.0012 s |
| Â | Transient | 61.98% | 61.98% | Â |
| Â | Speech with transient | 91.59% | 27.06% | Â |
Kernel | Speech | 88.18% | 66.95% | 0.2216 s |
| Â | Transient | 40.56% | 40.56% | Â |
| Â | Speech with transient | 86.87% | 72.98% | Â |
Proposed | Speech | 87.73% | 88.75% | 0.0093 s |
| Â | Transient | 81.54% | 79.35% | Â |
| Â | Speech with transient | 88.25% | 71.73% | Â |