Table 12.

Comparison of classification accuracy on CK+-O with existing models

Reference	Accuracy (%) on non-occluded faces	Accuracy (%) on occluded faces
Liu et al. (2018)	99.02	63.00
Wen et al. (2017)	76.05	–
Li et al. (2019b)	78.47	–
Zia et al. (2018)	92.00	–
Sun et al. (2019)	96.15	–
Poux et al. (2021)	91.3	88.8
Li et al. (2018a)	97.03	93.92
Dong et al. (2019)	93.00	90.30
Dapogny et al. (2018)	94.30	91.50
Proposed CES-CNN	96.64	95.76

Bold represents the better accuracy value of the proposed model