Audio-to-Visual Knowledge Distillation on RAVDESS
73.83AccuracyUCMKD
Evaluation results
| Method | Links | |
|---|---|---|
| UCMKDBackbone=ResNet-18, Batch size=64, Epoch=100, Optimizer=SGD, Learning rate=1e-2, Unpaired cross-modal setting=true2026.06 | 73.83 | |
| NORMBackbone=ResNet-18, Batch size=64, Epoch=100, Optimizer=SGD, Learning rate=1e-2, Unpaired cross-modal setting=true2026.06 | 64.97 | |
| REVIEWBackbone=ResNet-18, Batch size=64, Epoch=100, Optimizer=SGD, Learning rate=1e-2, Unpaired cross-modal setting=true2026.06 | 62.61 |