Skip to main content
. 2021 Nov 29;7:e736. doi: 10.7717/peerj-cs.736

Table 9. Comparison result of VGGML-CNN with some recent models on BU-3DFE.

Model Accuracy % No of classes Target
Fernandez et al. (2019) (FERAtt) 85.15 7 Expression only
Shao & Qian (2020) (MVFE-LightNet + Residual Convolution) 88.70 6 Expression only
Bao, Zhao & Chen (2020) (CNM) 80.63 6 Expression only
VGGML-CNN 98.01 6 Expression and intensity