Wednesday 24 August 2022

DALL-E 2 video

Hologram Files based on these AI images

I made a YouTube video of the best creations I made using DALL-E 2:




I've had access to DALL-E 2 for a week (since August 17th).

Other DALL-E 2 posts:


PROMPTS and PROGRESS

Below are the prompts and subsequent developments of the images:

01:

"black and white print of an artist working in the style of 
Leopoldo Mendez" --
I continued with the first image

I generated new images
based on the previous first image --
then I chose the second result (third image) to work with


I gave the image more volume
using Deep Dream Generator


"An artist working in his studio"  --
I chose the third result (fourth image)


"Artist in his studio"


Google Lens
Final result



02:


I uploaded this drawing from July 28, 2022,
and erased the head


I erased the face and asked DALL-E 2 to replace it with "a beautiful woman":

"a beautiful woman"  --
I chose the second result (third image) to continue



I then chose the first result (second image)



 I altered the first result from the image above,
in Deep Dream Generator


"a beautiful woman in a bikini"


Google Lens

I liked the first result the best


03:

I altered another one of my drawings:


I uploaded this drawing
from August 18, 2022



After uploading the drawing I erased most of the figure


"in the style of Egon Schiele" --
I chose the first result (second image) to work with


I chose the first result and applied the style from the image below using Deep Dream Generator:


Last year I created this fake Egon Schiele
using a VQGAN+CLIP program --
and I applied this style to the fake Egon Schiele
I made in DALL-E 2



This AI interpretation revisited the Komar & Melamid's project, when the two Russian artists took a survey of the qualities for the "most wanted" painting of each country, and then created paintings based on the results (YouTube explanation).


I entered the verbal description of America's Most Wanted Painting to see how DALL-E 2 would interpret it:


I chose the first image


First result, slightly altered by Deep Dream Generator




07:

"Tapir with six legs walking on a lake
in the style of unreal engine"  --
I chose the third image


Google Lens
I altered the third image
in Deep Dream Generator


Google Lens
I further edited the image


08:


"Three javelinas driving a spaceship
over Roswell, New Mexico"  --
I chose the fourth image


I asked for further variations of the last iamge from the above


Google Lens
I altered the first result (second image) from above
with Deep Dream Generator


The final edited alteration




09:

"Black and white image of two armadillos
walking past a cactus rendered in 3D"  --
I chose the second image

"Armadillos walking through the desert
during a volcanic eruption"


Google Lens
Altered in Deep Dream Generator
and further edited


10:


"Herd of tapirs and armadillos
charging out of a canyon | 795"  --
I chose the third image


Google Lens
Altered in Deep Dream Generator



11:

Google Lens
I first created this image in Photofunia (Concrete Jungle)


Then I "outpainted" the scene in DALL-E 2 to enlarge it, getting further and more distant views:

"City scene in Shibuya, Tokyo"  --
I chose the last result


Google Lens
Fake Shibuya, Tokyo




"Food cart in Shibuya, Tokyo" --
I chose the first result (second image)


Google Lens
Faker Shibuya, Tokyo



"Scene in Shibuya, Tokyo"  --
I chose the second result (third image)


Google Lens
Stepping back



"Tokyo as scene from space"  --
I chose the first result (second image)


Google Lens
Furthest view


Other AI Art Programs


Posted by at

No comments:

Post a Comment

Note: only a member of this blog may post a comment.

Subscribe to: Post Comments (Atom)