How is G(z) related to x in GAN proof?

4

1

In the proofs for the original GAN paper, it is written:

$$∫_x p_{data}(x) \log D(x)dx+∫_zp(z)\log(1−D(G(z)))dz =∫_xp_{data}(x)\log D(x)+p_G(x) \log(1−D(x))dx$$

I've seen some explanations asserting that the following equality is the key to understanding:

$$E_{z∼p_z(z)}log(1−D(G(z)))=E_{x∼p_G(x)}log(1−D(x))$$

which is a consequence of the LOTUS theorem and $x_g = g(z)$. Why is $x_g = g(z)$?

NNLearner

Posted 2019-06-15T23:16:55.513

Reputation: 63

in all formulas shown you never showed $x_g$ or $g(z)$. What do these refer to and where? – mshlis – 2019-06-17T20:14:09.867

Answers

1

It's not supposed to be derived from some equation. That is the basic premise under which GANs work. The output of the Generator $G(z)$ is fed as an input $x_g$ to the discriminator.

Djib2011

Posted 2019-06-15T23:16:55.513

Reputation: 2 624