1 Correction to: Circuits Systems Signal Process https://doi.org/10.1007/s00034-022-02008-4

The original version of this article is missing the below unnumbered equation to be available before equation (3),

Cycle-consistency loss: Advancing just the adversarial loss would not really ensure that the contextual detail of s and G(s) will be persistent. Since, the adversarial loss only explains about the variation among the generated (G(s)) and target domain (T) and doesn’t assist in preserving the contextual detail of s. In view of this one extra term namely cycle-consistency loss is added in CycleGAN, given as

$$\begin{aligned} {\mathcal {L}} _ {cyc} (G,F,S,T)&= {\mathbb {E}}_{s \sim p (S)} \left[ \Vert F(G(s)) - s \Vert _ {1} \right] \\&\quad + {\mathbb {E}}_ {t\sim p (T)} \left[ \Vert G(F(t)) - t \Vert _ {1} \right] \end{aligned}$$

The original article has been corrected.