This video is amazing! I usually don't watch long videos till the end because of my short attention span haha, this was the first video in months that I fully watched and I didn't even pause. I am in love with your way of explaining! Continue your work, I wish you all the best!
For example, if we provide 1000 images of ducks, can they all be retained by the model? What if we provide 100,000 more objects, each with 1000 images?
I would love to also see an explanation of how stable diffusion works 100% in Python, meaning with only the built-in modules. From watching this video, I might have figured out how to make a version that can generate 1280x1280 images under 1 second... even entirely on extremely slow modern computers because of the PIL.
Too bad the AIs refuse to give results that are perceived to be NSFW. It is creepy as hell. They can tell if a picture comes out to be nude when the prompt doesn't have any of those keywords. How do you make such high resolutions like the blue stained glass? Tools on the web are limited from 512 to 1152 pixels.
Ai is the end of Photography as we know it.. Soon you won't be able to tell what is Real and what's Fake. You will see a picture of a flawless Beautiful woman.. But in reality she is far from beautiful..
![The moment we stopped understanding AI [AlexNet]](http://i.ytimg.com/vi/UZDiGooFs54/mqdefault.jpg)
![The moment we stopped understanding AI [AlexNet]](/img/tr.png)
This deserves way more views.
How come such high quality video doesn’t have millions of views.
My thoughts exactly 😢
This video is amazing! I usually don't watch long videos till the end because of my short attention span haha, this was the first video in months that I fully watched and I didn't even pause. I am in love with your way of explaining! Continue your work, I wish you all the best!
we are among 5.3k that watch this fantastic video..
this video is so underrated! thanks for the explanation!
For example, if we provide 1000 images of ducks, can they all be retained by the model?
What if we provide 100,000 more objects, each with 1000 images?
great video, thanks. I guessed completely wrong on the image generators. Although I have used 2 of them many many times..... lol.
I would love to also see an explanation of how stable diffusion works 100% in Python, meaning with only the built-in modules. From watching this video, I might have figured out how to make a version that can generate 1280x1280 images under 1 second... even entirely on extremely slow modern computers because of the PIL.
Great! I found that information which searched for a long time
This is a really useful overview! Thank you. 🙂
Incredible video
Very good explained
Great job!!!
Next one can be about AI Generated videos :)
haha =)
Too bad the AIs refuse to give results that are perceived to be NSFW. It is creepy as hell. They can tell if a picture comes out to be nude when the prompt doesn't have any of those keywords. How do you make such high resolutions like the blue stained glass? Tools on the web are limited from 512 to 1152 pixels.
Ai is the end of Photography as we know it.. Soon you won't be able to tell what is Real and what's Fake. You will see a picture of a flawless Beautiful woman.. But in reality she is far from beautiful..