Wololo! VC leo umesafisha wanakijiji macho. Kindly take over from @nairobilay.
Halafu hio mashakura imeharibiwo na almonds. Nyama yenye iko na sukari? He’ll no!
There was a camera app that was developed hivi maajuzi that could undress anyone on the other side of the camera. It was immediately banned and deleted after its launch. Wapi @Deorro atuambie kinagaubaga
Unless anatuletea pirated download ya hio app acha akae tu huko diaspora ya ronga
@Deorro njoo na usitumane
I really miss him. His cryptic sense of humour is like no other. The grasp of lingua unmatched. Makes no apologies for his lavish lifestyle which he has truly worked hard for.
lemme try
That I guess used GAN - Generative adversarial Networks. This are just 2 neural networks pitted against each other where one acts as a generator while the other as a discriminator
Neural networks are just algorithms modelled on the human brain or nervous system to recognise patterns. just another application of Biomimicry. to create one you train it on a dataset then it will finally recognise the patterns.
Neural nets work in 3 ways: classification, clustering and regression.
Classification is where the neural et is given labelled datasets then it learns from those such that when give new data it can classify it based on the data it trained on. A perfect example is email spam systems. to create such you have a dataset containing two categories one legit emails while the other spam emails. the neural net then learns the patterns on each such that when a new email is sent it can categorise it on its own.
Clustering is where it’s given unlabeled data then it classifies the data on its own based on the patterns it learns. assuming you are given a list of players with stats of assist, goals, shots on target, tackles, clearances etc then told to find a way to know which ones are goalkeepers, defenders, mids and strikers. clustering algos will divide them in such clusters since strikers will have similar qualities, defenders same, goalkeepers same…
regression is simply finding the correlation. for example you want to predict stock prices. a perfect example is how `uber uses it to set prices during say rush hours. they can use that to predict which areas might have more user requests then sed drivers there and also set prices.
Now that you know the various ways of creating them, let’s go back to GAN, we said that they are 2 neural nets which compete against each other. One is called a generator and the other is called a discriminator. the work of the generator is to generate while the discriminator determines if it’s real or fake. Say you want to create a GAN that generates images of black women. You will get a dataset of images of black women then train the generator to create such while discriminator will be learning to tell if that image is for a fake or real black lady. the generator will be creating random images and for each image the discriminator will check if its real or fake. This will continue while the generator improves on’s own till it gets to a point the discriminator cannot recognise it as a fake. at that point, the generator ow creates real believable images of black women.
in the case of your app the owner simply downloaded lots of nude images then trained the GAN in such a way that it can generate the nude body based o the head given…okaysay you find the correlation between oblong shape of @Makonika s head with his body then do that for a million other people like akina @Cogito ergo sum and their dolichocephaly heads. the generator will learn to create a body from that while the discriminator will check if that’s a real human body or not. Train it to the point the discriminator cannot see it as a fake. next when you use your cam to take a pic of @Mrs4thletter it ca generate a nude body based on what it learnt…
…
…
nimechoka, wacheni iende nipike omena, lakini I hope you guys learnt how that shit works. for those who did not understand I am not a good teacher but one day we might talk about machine learnning, deep learning, ai stuff inn depth
Weka spoiler, nugu wewe!
Chieth.