Multipose Face Recognition-Based Combined Adaptive Deep Learning Vector Quantization

Table 7

Comparison between ADLVQ and CADLVQ.

	ADLVQ	CADLVQ
Hyper-parameters	Fixed	Variable
Adaptation	Adaptation in procedure stage	Adaptation in both procedure stage and hyper-parameter inputs
Complexity	High	Very high
Maximum memory size	1.5 GB	2 GB
Average accuracy	96.62%	98.53%