A New Way to look at GANs

This is my first post here!
My name is Marco, I’m 20 years old (21 very soon!) and I’m from Italy. Last year in February I decided to educate myself on Deep Learning; I’ve never done any programming in high school (good old Latin instead) so I had to start from the basics.

After getting to know Python, I followed some online courses on DL, and the fast.ai course was the best one for me: right to the point and super practical (I’m doing all of this while studying Energy Engineering and I don’t have much time on my hands).

Fast forward to today: I got very interested in GANs and I wanted to share what I found experimenting (and completely losing my head sometimes) with them.
In short, using a Siamese Discriminator and making the Discriminator output a vector instead of a single value can be useful to some tasks.

This is my article (it’s my first one ever, some feedback is very appreciated!):

I’ve tried looking for some similar experiments but I couldn’t find them, though I’m quite sure there are. So if you know something, please let me know!

Forgot to mention, all the testings were done in Colab (with sometimes huge waiting times).