Ear Recognition Based on Gabor Features and KFDA

Table 3

Rank-1 recognition rate of the proposed method when adopting different kernel functions.

Kernel	𝐾(𝑥,𝑦)	Parameters in kernel functions	Rank-1 recognition rate
			USTB dataset 3	UND dataset
Linear	𝐾(𝑥,𝑥𝑖)=(𝑥⋅𝑥𝑖)	—	75.44%	74.22%
Polynomial	𝐾(𝑥,𝑥𝑖)=𝑎𝑥⋅𝑥𝑖+𝑏^𝑑	𝑎=0.001, 𝑏=1, 𝑑=2	89.11%	87.33%
RBF	𝐾(𝑥,𝑥𝑖)=exp−𝛾𝑥−𝑥𝑖^2	𝛾=0.0001	96.46%	94%
Cosine	𝑘(𝑥,𝑦)=𝑘(𝑥,𝑦)𝑘(𝑥,𝑥)𝑘(𝑦,𝑦); 𝑘(𝑥,𝑦)=𝑎𝑥⋅𝑦+𝑏^𝑑	𝑎=0.001, 𝑏=1, 𝑑=2	92.41%	90.22%