Research Article
Facial Expression Recognition of Instructor Using Deep Features and Extreme Learning Machine
Table 3
Results of deep model layers on standard datasets (CK, Jaffe, and FER2013) using validation schemes across RELM classifier.
| | Dataset | Network | Layer | Split (70–30%) | 10-fold | LOAO | LOSO |
| | Jaffe | DenseNet201 | conv4_block9_1_bn | 96.8 | 92.44 | 93.4 | 94.55 | | AlexNet | drop7 | 91.67 | 86.03 | 84.01 | 85.91 |
| | CK | DenseNet201 | conv4_block9_1_bn | 86.59 | 80.5 | 81.59 | 81.99 | | AlexNet | drop7 | 81.67 | 86.03 | 84.01 | 85.91 |
| | FER2013 | DenseNet201 | conv4_block9_1_bn | 62.74 | — | — | — | | AlexNet | drop7 | 45.91 | — | — | — |
|
|
