Voice Activity Detection on LibriVAD-NonConcat small-scale (train and test)
98.9VAD Performance (Nature)GFCC-ViT
Evaluation results
| Method | Links | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| GFCC-ViTSNR=20, Feature=GFCC, Architecture=ViT2025.12 | 98.9 | 98.72 | 98.59 | 98.8 | 98.87 | 98.62 | 97.83 | 98.72 | 98.83 | — | |
| MFCC-ViTSNR=20, Feature=MFCC, Architecture=ViT2025.12 | 98.88 | 98.64 | 98.49 | 98.81 | 98.75 | 98.67 | 97.08 | 98.48 | 98.9 | — | |
| MFCC-ViTSNR=15, Feature=MFCC, Architecture=ViT2025.12 | 98.76 | 98.62 | 98.14 | 98.7 | 98.78 | 98.49 | 95.7 | 98.23 | 98.8 | — | |
| GFCC-ViTSNR=15, Feature=GFCC, Architecture=ViT2025.12 | 98.68 | 98.58 | 98.25 | 98.77 | 98.75 | 98.38 | 96.97 | 98.47 | 98.75 | — | |
| MFCC-ViTSNR=10, Feature=MFCC, Architecture=ViT2025.12 | 98.49 | 98.19 | 97.74 | 98.44 | 98.6 | 98.01 | 92.75 | 97.88 | 98.57 | — | |
| GFCC-ViTSNR=10, Feature=GFCC, Architecture=ViT2025.12 | 98.37 | 98.34 | 97.68 | 98.49 | 98.65 | 98.06 | 95.05 | 97.86 | 98.77 | — | |
| MFCC-ViTSNR=5, Feature=MFCC, Architecture=ViT2025.12 | 98.15 | 97.62 | 96.68 | 97.83 | 98.31 | 97.2 | 86.62 | 97.12 | 98.4 | — | |
| MFCC-ViTSNR=Avg, Feature=MFCC, Architecture=ViT2025.12 | 97.87 | 97.31 | 95.38 | 97.58 | 98.2 | 96.38 | 84.69 | 96.01 | 98.23 | 0.9574 | |
| GFCC-ViTSNR=5, Feature=GFCC, Architecture=ViT2025.12 | 97.84 | 97.79 | 96.76 | 98.02 | 98.35 | 97.34 | 90.41 | 97.24 | 98.51 | — | |
| GFCC-ViTSNR=Avg, Feature=GFCC, Architecture=ViT2025.12 | 97.73 | 97.52 | 95.61 | 97.55 | 98.3 | 96.35 | 86.92 | 96.63 | 98.31 | 0.961 | |
| MFCC-ViTSNR=0, Feature=MFCC, Architecture=ViT2025.12 | 97.41 | 96.47 | 94.27 | 96.9 | 97.83 | 94.95 | 74.75 | 94.95 | 97.8 | — | |
| GFCC-ViTSNR=0, Feature=GFCC, Architecture=ViT2025.12 | 97.12 | 96.95 | 94.44 | 97.02 | 97.84 | 95.26 | 77.68 | 95.83 | 97.98 | — | |
| raw-CLDNNSNR=20, Feature=raw, Architecture=CLDNN2025.12 | 95.62 | 95.81 | 94.51 | 95.9 | 96.6 | 94.76 | 91.37 | 94.32 | 95.53 | — | |
| MFCC-ViTSNR=-5, Feature=MFCC, Architecture=ViT2025.12 | 95.52 | 94.34 | 86.94 | 94.8 | 96.95 | 90.94 | 61.23 | 89.37 | 96.89 | — | |
| GFCC-ViTSNR=-5, Feature=GFCC, Architecture=ViT2025.12 | 95.45 | 94.75 | 87.93 | 94.22 | 97.36 | 90.44 | 63.55 | 91.67 | 96.99 | — | |
| raw-CLDNNSNR=15, Feature=raw, Architecture=CLDNN2025.12 | 94.79 | 95.05 | 92.81 | 95.02 | 96.31 | 93.09 | 86.6 | 92.57 | 93.9 | — | |
| MFCC-bDNNSNR=20, Feature=MFCC, Architecture=bDNN2025.12 | 94.66 | 95.18 | 95.5 | 95.24 | 95.92 | 95.19 | 74.73 | 96.36 | 96.39 | — | |
| MFCC-bDNNSNR=15, Feature=MFCC, Architecture=bDNN2025.12 | 94.18 | 95.07 | 94.3 | 94.68 | 95.93 | 94.39 | 69 | 95.43 | 96.27 | — | |
| MFCC-bDNNSNR=10, Feature=MFCC, Architecture=bDNN2025.12 | 93.33 | 94.74 | 92.1 | 93.51 | 95.77 | 92.92 | 62.44 | 93.52 | 95.92 | — | |
| raw-CLDNNSNR=10, Feature=raw, Architecture=CLDNN2025.12 | 93.06 | 93.86 | 90.19 | 93.45 | 95.77 | 90.44 | 79.13 | 89.6 | 91.05 | — | |
| MFCC-bDNNSNR=5, Feature=MFCC, Architecture=bDNN2025.12 | 91.73 | 93.88 | 88 | 91.34 | 95.21 | 90.29 | 55.41 | 89.92 | 95.13 | — | |
| GFCC-bDNNSNR=20, Feature=GFCC, Architecture=bDNN2025.12 | 91.47 | 91.38 | 91.82 | 91.12 | 92.86 | 91.32 | 65.56 | 94.69 | 94.33 | — | |
| GFCC-bDNNSNR=15, Feature=GFCC, Architecture=bDNN2025.12 | 91.42 | 91.88 | 90.96 | 90.91 | 93.28 | 91.08 | 62.38 | 94.02 | 94.67 | — | |
| MFCC-bDNNSNR=Avg, Feature=MFCC, Architecture=bDNN2025.12 | 91.16 | 93.4 | 87.24 | 90.83 | 94.78 | 89.7 | 59.41 | 88.83 | 94.69 | 0.8785 | |
| GFCC-bDNNSNR=10, Feature=GFCC, Architecture=bDNN2025.12 | 90.81 | 92.08 | 88.98 | 90.21 | 93.46 | 90.01 | 58.24 | 92.18 | 94.61 | — | |
| raw-CLDNNSNR=5, Feature=raw, Architecture=CLDNN2025.12 | 90.79 | 92.18 | 86.24 | 90.94 | 94.81 | 86.83 | 69.69 | 84.79 | 87.18 | — | |
| raw-CLDNNSNR=Avg, Feature=raw, Architecture=CLDNN2025.12 | 90.65 | 91.89 | 85.77 | 90.62 | 94.51 | 86.79 | 73.23 | 84.13 | 87.93 | 0.8728 | |
| GFCC-bDNNSNR=5, Feature=GFCC, Architecture=bDNN2025.12 | 89.43 | 91.74 | 85.03 | 88.37 | 93.1 | 87.65 | 53.39 | 88.53 | 94.02 | — | |
| MFCC-bDNNSNR=0, Feature=MFCC, Architecture=bDNN2025.12 | 88.91 | 92.13 | 81.21 | 87.78 | 93.99 | 85.91 | 49.15 | 83.64 | 93.57 | — | |
| GFCC-bDNNSNR=Avg, Feature=GFCC, Architecture=bDNN2025.12 | 88.81 | 91.06 | 84.21 | 87.6 | 92.49 | 86.83 | 55.8 | 87.21 | 93.42 | 0.8527 | |
| raw-CLDNNSNR=0, Feature=raw, Architecture=CLDNN2025.12 | 87.38 | 89.47 | 79.83 | 87.2 | 93.1 | 81.64 | 60.08 | 77.19 | 82.65 | — | |
| GFCC-bDNNSNR=0, Feature=GFCC, Architecture=bDNN2025.12 | 86.91 | 90.62 | 78.55 | 85 | 92.01 | 83.5 | 48.78 | 81.99 | 92.68 | — | |
| MFCC-bDNNSNR=-5, Feature=MFCC, Architecture=bDNN2025.12 | 84.19 | 89.41 | 72.33 | 82.45 | 91.88 | 79.51 | 45.78 | 74.13 | 90.88 | — | |
| GFCC-bDNNSNR=-5, Feature=GFCC, Architecture=bDNN2025.12 | 82.83 | 88.69 | 69.96 | 80.02 | 90.25 | 77.42 | 46.47 | 71.89 | 90.21 | — | |
| raw-CLDNNSNR=-5, Feature=raw, Architecture=CLDNN2025.12 | 82.29 | 85.01 | 71.04 | 81.22 | 90.49 | 74.02 | 52.54 | 66.35 | 77.3 | — |