An Overview of Multimodal Biometrics Using the Face and Ear

Table 8

Multimodal recognition accuracy (%) versus random pixel corruption on MD III.

Corruption	0%	20%	40%	60%	80%	100%
(a) Face corruption
MNN	75.129	71.595	67.242	59.021	47.531	30.227
MNFL	78.133	72.374	67.909	63.194	47.114	30.964
MSVM	87.717	87.091	85.13	78.717	64.946	45.194
MCCA	91.946	78.008	51.774	26.151	12.102	6.232
MSRCef	98.345	97.093	95.006	89.22	77.813	59.452
MSRCs	98.873	97.719	95.187	90.75	84.226	75.031
MRSC	98.512	97.148	95.41	91.084	90.291	87.69
AMRSC	98.998	98.317	95.284	93.337	92.475	92.141
(b) Ear corruption
MNN	75.129	69.245	47.976	27.862	19.224	15.579
MNFL	78.133	72.138	48.004	28.391	20.573	17.012
MSVM	87.717	75.796	31.785	16.247	11.156	8.624
MCCA	91.946	83.642	73.306	63.722	54.013	48.157
MSRCef	98.345	97.454	94.728	89.901	76.603	63.903
MSRCs	98.873	98.025	94.505	89.025	80.581	70.135
MRSC	98.512	97.691	96.634	90.958	88.079	89.428
AMRSC	98.998	98.859	97.26	96.467	96.133	95.549