[Retracted] Feature Learning-Based Generative Adversarial Network Data Augmentation for Class-Based Few-Shot Learning

Table 4

Random 1,000 samples classification accuracy and GQI.

Data sets	Accuracy (%)	GQI	LS
MNIST (full)	99.13	1	0.886
VAE	95.92	0.9677	0,497
CGAN	96.63	0.9748	0.698
DAGAN	96.72	0.9756	0.738
f-DAGAN	98.93	0.9979	0.746