TABLE 3.

The mean classification accuracy of each AFERS on faces of the same sex and expression.

Expression	Dataset	FaceReader				DeepFace
		Nimages	Male	Nimages	Female	Nimages	Male	Nimages	Female
Angry	Chen’s	12	0.833	7	1.000	12	0.833	7	0.571
	Tu’s	8	0.250	7	0.429	8	0.500	7	0.429
	KDEF	19	1.000	12	0.917	19	0.842	12	1.000
	RaFD	14	1.000	11	1.000	14	0.857	11	0.727
Happy	Chen’s	106	0.962	108	0.972	100	0.870	107	0.972
	Tu’s	27	0.926	27	0.889	27	0.778	27	0.704
	KDEF	28	1.000	31	1.000	28	1.000	31	1.000
	RaFD	20	1.000	19	1.000	20	1.000	19	1.000
Sad	Chen’s	11	0.455	23	0.826	11	0.273	23	0.522
	Tu’s	2	0.000	10	0.300	2	0.500	10	0.600
	KDEF	9	1.000	11	1.000	9	0.778	11	0.818
	RaFD	10	1.000	13	1.000	10	0.800	13	0.769
Surprised	Chen’s	35	0.914	43	0.930	33	0.758	43	0.698
	Tu’s	21	0.952	18	0.944	21	0.571	18	0.444
	KDEF	5	1.000	12	1.000	5	1.000	12	0.833
	RaFD	14	1.000	13	1.000	14	0.643	13	0.462
Total	Chen’s	164	0.909	181	0.945	156	0.801	180	0.833
	Tu’s	58	0.810	62	0.758	58	0.655	62	0.581
	KDEF	61	1.000	66	0.985	61	0.918	66	0.939
	RaFD	58	1.000	56	1.000	58	0.845	56	0.768