Skip to main content
. 2013 Dec 5;13(12):16682–16713. doi: 10.3390/s131216682

Table 8.

The recognition rate of the HL-FER at the first level for recognizing the expressions category such as Lips-Based, Lips-Eyes-Based, and Lips-Eyes-Forehead-Based on the AT&T dataset (Unit: %).

Expressions Category Recognition Rate
Lips-Based 87
Lips-Eyes-Based 84
Lips-Eyes-Forehead-Based 82
Average 84.3