Valence Estimation on AffectNet (test)
0.62CCCVGG-FACE trained using the proposed approach
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| VGG-FACE trained using the proposed approachNetworks=VGG-FACE trained using the proposed approach2018.11 | 0.62 | 0.66 | 0.78 | 0.14 | |
| AlexNetNetworks=AlexNet2018.11 | 0.6 | 0.66 | 0.74 | 0.14 | |
| VGG-FACE trained using GanimationNetworks=VGG-FACE trained using Ganimation2018.11 | 0.56 | 0.59 | 0.74 | 0.15 | |
| VGG-FACE trained using StarGANNetworks=VGG-FACE trained using StarGAN2018.11 | 0.55 | 0.58 | 0.74 | 0.17 | |
| VGG-FACE baselineNetworks=the VGG-FACE baseline2018.11 | 0.5 | 0.54 | 0.65 | 0.19 |