Facial Expression Recognition on SZU-EmoDage
90.91AccuracyCS3D
Evaluation Results
| Method | Links | |
|---|---|---|
| CS3Doptimizer=Adam, learning rate=1e-4, batch size=162025.12 | 90.91 | |
| C3D3D convolutional layers=5, fully connected layers=2, optimizer=Adam, learning rate=1e-4, batch size=162025.12 | 79.45 | |
| RNNrecurrent layers=2, hidden units=128, optimizer=Adam, learning rate=1e-4, batch size=162025.12 | 33.33 | |
| LSTMrecurrent layers=2, hidden units=128, optimizer=Adam, learning rate=1e-4, batch size=162025.12 | 29.55 | |
| Transformerencoder layers=2, attention heads=4, model dimension=256, optimizer=Adam, learning rate=1e-4, batch size=162025.12 | 28.03 |