Classification on Google commands
98.59AccuracyFMix
Evaluation Results
| Method | Links | |
|---|---|---|
| FMixModel=ResNet, alpha=12020.02 | 98.59 | |
| CutMixModel=ResNet, alpha=0.22020.02 | 98.48 | |
| MixUpModel=ResNet, alpha=12020.02 | 98.46 | |
| CutMixModel=ResNet, alpha=12020.02 | 98.46 | |
| FMixModel=ResNet, alpha=0.22020.02 | 98.44 | |
| MixUpModel=ResNet, alpha=0.22020.02 | 98.31 | |
| BaselineModel=ResNet, alpha=0.22020.02 | 97.69 | |
| LOES (3)Model=Wav2Vec 2.0 (Frozen), k=32026.05 | 92.6 | |
| LOES (4)Model=Wav2Vec 2.0 (Frozen), k=42026.05 | 92.51 | |
| LOES (2)Model=Wav2Vec 2.0 (Frozen), k=22026.05 | 92.41 | |
| Last LayerModel=Wav2Vec 2.0 (Frozen)2026.05 | 85.15 | |
| MetaMAEPretrain data=LibriSpeech, Protocol=Linear evaluation2023.10 | 46.8 | |
| ShEDPretrain data=LibriSpeech, Protocol=Linear evaluation2023.10 | 20.7 | |
| e-MixPretrain data=LibriSpeech, Protocol=Linear evaluation2023.10 | 19.2 | |
| CapriPretrain data=LibriSpeech, Protocol=Linear evaluation2023.10 | 13.7 | |
| MAEPretrain data=LibriSpeech, Protocol=Linear evaluation2023.10 | 9.5 | |
| BaselinePretrain data=LibriSpeech, Protocol=Linear evaluation2023.10 | 4.9 |