FLGC-Fusion GAN: An Enhanced Fusion GAN Model by Importing Fully Learnable Group Convolution

Table 3

Run-time comparison of six methods on the TNO datasets.

Method	TNO
CVT	1.46
DTCWT	3.32 × 10^−1
GFF	3.18 × 10^−1
GTF	4.82
Fusion GAN	4.61 × 10^−2
FLGC-Fusion GAN	3.07 × 10^−2

The CVT, DTCWT, GFF, GTF, Fusion GAN, and FLGC-Fusion GAN are performed on GPU while all the others are performed on CPU. Each value denotes the mean of the run time of a certain method on a dataset (unit: second).