Keyword Spotting on GSC1-12 (test)
98.25KWS AccuracySRE + BiLSTM multi
Evaluation Results
| Method | Links | |
|---|---|---|
| SRE + BiLSTM multiTrain Dataset=VoxCeleb2 + GSC1 (12), Training Protocol=multi-task, Downstream Head=BiLSTM2021.10 | 98.25 | |
| SRE + Linear multiTrain Dataset=VoxCeleb2 + GSC1 (12), Training Protocol=multi-task, Downstream Head=Linear2021.10 | 98.23 | |
| SRE + CNN multiTrain Dataset=VoxCeleb2 + GSC1 (12), Training Protocol=multi-task, Downstream Head=CNN2021.10 | 98.21 | |
| SRE + Linear multiTrain Dataset=VoxCeleb1 + GSC1 (12), Training Protocol=multi-task, Downstream Head=Linear2021.10 | 98.19 | |
| SRE + BiLSTM multiTrain Dataset=VoxCeleb1 + GSC1 (12), Training Protocol=multi-task, Downstream Head=BiLSTM2021.10 | 98.19 | |
| SRE + CNN multiTrain Dataset=VoxCeleb1 + GSC1 (12), Training Protocol=multi-task, Downstream Head=CNN2021.10 | 98.11 | |
| SRE + Linear singleTrain Dataset=GSC1 (12), Training Protocol=single-task, Downstream Head=Linear2021.10 | 98.1 | |
| BC-ResNet-8-KWSTrain Dataset=GSC1 (12), Backbone=ResNet-82021.10 | 98 | |
| Wav2KWSTrain Dataset=GSC1 (12), Backbone=Wav2Vec2021.10 | 97.9 | |
| SRE (Frozen) + LinearTrain Dataset=VoxCeleb1 + GSC1 (12), Backbone State=frozen, Training Protocol=multi-task, Downstream Head=Linear2021.10 | 90.33 | |
| SRE (Random) + LinearTrain Dataset=VoxCeleb1 + GSC1 (12), Backbone State=randomly initialized, Training Protocol=multi-task, Downstream Head=Linear2021.10 | 56.58 |