AI Image requests - Now generate her giving birth.

Do you mean a guy dressed as Barbie in a mugshot?
Or a guy photographing Barbie in the mugshot?
Kiwis in the mugshot?
Kiwis taking the mugshot?

Details please and I’d love to try.
oh sorry late reply, what i mean is a guy doing the barbie mugshot meme with kiwi birds beside him. images (99).jpeg
this is the image reference
 
Can someone do Larry Potterfield holding a katana? I need a new profile picture. Or else Donald Trump and Joe Biden in a fist fight.
 
oh sorry late reply, what i mean is a guy doing the barbie mugshot meme with kiwi birds beside him. View attachment 5355392
this is the image .
3EFF1062-D0A5-4F24-A00A-61E908035C7F.jpeg
Will accept constructive criticism. I love making things like this but sometimes it’s hard to put across in words, the end result required. Be sure to get back to me if I’ve got things wrong or you feel it could be improved. This is all a steep learning curve for me and I appreciate the opportunities to practice on tasks like this.
 
Last edited:
Can someone do Larry Potterfield holding a katana? I need a new profile picture. Or else Donald Trump and Joe Biden in a fist fight.
What do you think? Again….constructive criticism always welcome. I’m a boomer in spirit and trying to learn. I have a few different shop programs that work well together, but I’m still learning how to get the best results. Combining photoshooping and AI can lead to some fun effects. Some look great. Some look terrible.
As I said. I’m open to suggestions and criticism. At the moment I’m finding that a complete mix of real photos, AI interpretation and shopping can create a vast array of effects .052BEB3A-4EC6-4A32-B853-044E7E11B6CB.jpeg
This was done by feeding real pics into AI and then taking the result back to photoshop to tidy up, although doing these steps in different orders can produce better results. I’m Finding a steep learning curve.

I love getting assignments from this thread! You’re all funny Frens, and I’m learning on the job creating some bizarre images. One things for sure. You get the best results combining real images with AI and then manipulating the AI……and mixing back with real images.

Edit. Had a try at the Larry Potterfield Katana guy……first attempts….
21F28914-E04E-4CF3-BE11-338FCCE43212.jpeg4CC342D9-2026-4A18-91E1-39ACABDD2AFA.jpeg8992EBDE-E868-417B-B4A9-8E6E42B33244.jpeg
I think he looks way too happy. Again…constructive criticism is needed, I feel. I tried. I’m not sure I entirely hit it out of the park, but I’m learning.
 
Last edited:
Can you ask AI to spookify my pfp for halloween?

Like this?


BE6440DD-53A5-4E35-8A5F-4A9725BAAA8D.jpeg

Sorry, I lost the originals. These are the ones, no?


Ok. I had a go….. I’m not kidding when I say that “The Safety Bill” may have impacted my AI programs. I am spending ages trying to get some cool images for you, and they’re pretty tame…..I was finding this hard.
I’m trying.A01FE71B-D57C-4E7C-B38A-4ED2F43EB831.jpeg0AC7F2CB-6E49-4A05-9DB1-B8621D01A91E.jpeg

0EC34A3B-8B3E-4AF6-ACEB-3900373D72FE.jpeg0EC34A3B-8B3E-4AF6-ACEB-3900373D72FE.jpeg

Sorry. I’ll have another go when I haven’t consumed half a handle of delicious Sicilian Lemon Rum and I’m more sure I’m producing some interesting. Constructive criticism gratefully accepted, fren.
 

Attachments

  • 3C265619-A1F0-44D8-9DF7-B3D38FBC8533.jpeg
    3C265619-A1F0-44D8-9DF7-B3D38FBC8533.jpeg
    140.7 KB · Views: 10
  • Like
Reactions: Pedophobe
Like this?


View attachment 5360395

Sorry, I lost the originals. These are the ones, no?


Ok. I had a go….. I’m not kidding when I say that “The Safety Bill” may have impacted my AI programs. I am spending ages trying to get some cool images for you, and they’re pretty tame…..I was finding this hard.
I’m trying.View attachment 5360321View attachment 5360322

View attachment 5360383View attachment 5360383

Sorry. I’ll have another go when I haven’t consumed half a handle of delicious Sicilian Lemon Rum and I’m more sure I’m producing some interesting. Constructive criticism gratefully accepted, fren.
Thank you sir, excellent work, I have made my choice.
 
  • Like
Reactions: Suspicious Activity
I wanted to request if it was possible to make a hi-res version of these Mario 64 skyboxes. It's hard to just resize them without the quality dropping to shit (Since they are 640x480 images).

Just love how they look, wonder if AI-ing can help.

bints.jpglll.jpgtotwc.jpghbm.jpgwdw.jpgccc.jpg
 
  • Thunk-Provoking
Reactions: The Last Stand
I have two challenges if anyone wants a bash

1. Bog-standard poster

2. mixing imagines of vinyls melting in an oven
1695593709684.png
with this one
1695593814375.png
except the oven has swastikas for knobs/dials/handles

Cheers
 
This was done by feeding real pics into AI and then taking the result back to photoshop to tidy up, although doing these steps in different orders can produce better results. I’m Finding a steep learning curve.
Looks good. I love how it looks like Trump's hair is bolted on.
think he looks way too happy. Again…constructive criticism is needed, I feel. I tried. I’m not sure I entirely hit it out of the park, but I’m learning.
That's actually pretty accurate to how he is is in real life; he's always smiling. Thanks, I'm definitely gonna use the middle one at some point.
 
As for your question about video games covers and templates, I have no idea what this really means. So, I’d take an image of the original cover and then…..what? I don’t think I have access to ‘default templates’. In my shoops and AI feeds I rely on my own basic art skills and then knowledge of manipulating the AI the right way……I’m not sure where anything ‘default’ comes in…..but, I’m not “All-Knowing”….and so require a little extra help here.
I'm sorry. Let me try to explain it better. Video game cover: superimpose your AI generated image into a template like this:

1695788289369.png


To make video game cover art.

Example: Could you do Joan Crawford with her arms crossed holding knives with a title like "Slasher Dame" in the middle with that Xbox template?

If that makes sense.

Edit: if you need a guide: try this for size.

1695788733525.png

If AI CAN do what I'm asking with some editing.
 
Last edited:
I wanted to request if it was possible to make a hi-res version of these Mario 64 skyboxes. It's hard to just resize them without the quality dropping to shit (Since they are 640x480 images).

Just love how they look, wonder if AI-ing can help.

View attachment 5360461View attachment 5360464View attachment 5360466View attachment 5360471View attachment 5360472View attachment 5360480
I ran one through ‘Dream’ using a mode and description to replicate exactly under highest quality but I honestly don’t know if it improves the size or definition of the image at all….
FWIW, here it is.
7ADE5FB9-559D-44D7-8D74-94422E534E1F.jpeg
Thumbnail

7ADE5FB9-559D-44D7-8D74-94422E534E1F.jpeg


[/SPOLIER]
If this manoeuvre is useful to you, I’ll do all the others. I’m not very tech-y and I’m mobile so I do what I can. If the program does anything to increase the definition I wouldn’t know how to check it properly. They don’t look bad to my eye, but you can let me know.

(Sorry, I uploaded the results with the watermark. I am allowed to take them off and usually do, so that’s not a thing.)
 
  • Feels
Reactions: The Last Stand
Sorry for the triple post! I was hoping to end up with something good. I will gladly accept all the negrates.
You're way better than many of us can ever dream about.

Do you have the ability to extend existing photos with AI?

1695844809419.png

Could you add birds in the foreground?
 
I'm sorry. Let me try to explain it better. Video game cover: superimpose your AI generated image into a template like this:

View attachment 5365621

To make video game cover art.

Example: Could you do Joan Crawford with her arms crossed holding knives with a title like "Slasher Dame" in the middle with that Xbox template?

If that makes sense.

Edit: if you need a guide: try this for size.

View attachment 5365644

If AI CAN do what I'm asking with some editing.
I tried…
D0A85D60-DEFD-4886-83E5-2CB4BE7BDE6C.jpeg
 
  • Winner
Reactions: The Last Stand
Beautiful. I'll take two copies. I should've added an ESRB rating for the bottom right corner. It's way better than many fan art cover arts going around.

Actually, let's have some fun.

View attachment 5366947

Have the cover be a Monroe esque woman dancing in a ballroom, stylized as a painting. Name it whatever you want. Go nuts.
Like this?

AFF2AAD5-4182-48C2-9564-C989D38AB210.jpeg
277AF864-7F4C-4FFB-9EF8-B9B61DDB4763.jpegC4C40A19-E59B-40FE-8837-7B66E6C0DC40.jpeg1D5A1B86-3104-459B-8D62-9448EC567244.jpeg
 
  • Winner
Reactions: The Last Stand
Exactly that. It is hard to make a convincing cover of anything.
Now with Wii waggle and slightly updated graphics.

I see that you're having fun lately with the AI tool for others.
I ran one through ‘Dream’ using a mode and description to replicate exactly under highest quality but I honestly don’t know if it improves the size or definition of the image at all….
FWIW, here it is.
7ADE5FB9-559D-44D7-8D74-94422E534E1F.jpeg
How hard was it to upscale an existing image? What was your process with merging the PNG cover with something else?


Here's something that may not involve AI, but I'll let you decide. Probably PhotoShop.

You know how Oreo has their branding as "Milk's Favorite Cookie" on the packaging?

1695847885303.png

I want something like this, but have it say "The User We Love to Hate" or "Kiwi Farms' Favorite Cookie" substituted for "Milk's Favorite Cookie." If possible, have the red triangle say something else. A parody of sorts for me to have on my profile.

I think it'd be easier for you if you can make it from scratch inspired as how Oreo does their packaging. Could it be in a dimension to where I could use it as a banner? Horizontal.


Assets if you need them:

1695848066114.png1695848125280.png
 
You're way better than many of us can ever dream about.

Do you have the ability to extend existing photos with AI?

View attachment 5366910

Could you add birds in the foreground?
Extend existing photographs? Yes, but the more you give it room to play, the more details of the original image you'll end up losing. Don't give it enough room, and the artifacts of extending the image are very visible. A mixture of image generation and photoshop would be the ideal way to go.

First steps would be removing everything that isn't the image itself (via cropping; better we do it ourselves then feed the cropped image to the AI), then changing the dimensions.
00004-3762526762.jpg00005-3211154676.jpg
Not quite the right prompt, and too much denoising strength.
00011-2504720825.jpg
Not enough denoising strength; see the oddities on the edges? I told it to fill with colors, but not enough oomph to get rid of that.
00012-2000858497.jpg
Better.
00015-3753201463.jpg00017-1335425984.jpg00018-1335425985.jpg
So we're losing quite a bit of detail of the original image and it can't do firearms properly, but we are getting something not unlike the original image but with different dimensions. Either photoshop to the rescue, or keep making them until you get one close enough to make hires.
photorealistic, woman firing a rifle, conservative clothing, birds, monochrome, plains
Negative prompt: BadDream, verybadimagenegative_v1.3, easynegative, out of focus, empty areas, white, empty, blank, anime, horizontal lines, nsfw, text, caption, overlay, gui
Steps: 55, Sampler: DPM++ SDE Karras, CFG scale: 9, Seed: 1335425984, Size: 768x432, Model hash: 879db523c3, Model: dreamshaper_8, Denoising strength: 0.55, Clip skip: 2, Mask blur: 4
Used embeddings: BadDream [48d0], verybadimagenegative_v1.3 [89de], easynegative [119b]
 
  • Informative
Reactions: The Last Stand
00011-2504720825.jpg
Not enough denoising strength; see the oddities on the edges? I told it to fill with colors, but not enough oomph to get rid of that.
Oh yes, I definitely see how the algorithm struggled with generating imagery with little room to expand.
Extend existing photographs? Yes, but the more you give it room to play, the more details of the original image you'll end up losing. Don't give it enough room, and the artifacts of extending the image are very visible. A mixture of image generation and photoshop would be the ideal way to go.
What program did you use to achieve that effect?

I assume with something simple like a skyline or a backyard, it can be sophisticated enough to be unnoticeable to the naked eye.

1695849976759.png

How can AI extend THIS photograph?
 
Back