Visual-to-Audio Knowledge Distillation on CREMA-D
66.67AccuracyUCMKD
Evaluation Results
| Method | Links | |
|---|---|---|
| UCMKDBackbone=ResNet-18, Batch size=64, Epoch=100, Optimizer=SGD, Learning rate=1e-2, Unpaired cross-modal setting=true2026.06 | 66.67 | |
| REVIEWBackbone=ResNet-18, Batch size=64, Epoch=100, Optimizer=SGD, Learning rate=1e-2, Unpaired cross-modal setting=true2026.06 | 60.62 | |
| NORMBackbone=ResNet-18, Batch size=64, Epoch=100, Optimizer=SGD, Learning rate=1e-2, Unpaired cross-modal setting=true2026.06 | 60.48 |