55
submitted 6 months ago* (last edited 6 months ago) by j4k3@lemmy.world to c/imageai@sh.itjust.works

The Up Side Down

I want to see the most impressive images of an upside down reality. From objects, to landscapes, or anything in between, what kinds of inceptions can you create.

extra creditI've been trying to create an O'Neill Cylinder space habitat interior for a week now. Major kudos if you can can defy gravity with one of these. If you don't know what an O'Neill cylinder is, watch this 3min render: https://www.youtube.com/watch?v=Z2d_0l5ycRM or here is the wiki

Rules

  • Follow the community’s rules above all else
  • One comment and image per user
  • Embed image directly in the post (no external link)
  • Workflow/Prompt sharing encouraged (we’re all here for fun and learning)
  • Posts that are tied will both get the points
  • The challenge runs for 7 days from now on
  • Down votes will not be counted

Scores

At the end of the challenge each post will be scored:

  • Most upvoted: +3 points
  • Second most upvoted: +2 pointS
  • Third most upvoted: +1 point
  • OP’s favorite: +1 point
  • Most unconformable: +1 point
  • Last two entries (to compensate for less time to vote): +1 point
  • Prompt and workflow included: +1 point

The winner gets to pick next theme! Have fun everyone!

all 34 comments
sorted by: hot top controversial new old
[-] cloudless@lemmy.cafe 7 points 6 months ago

I have no idea what OP is asking, so here is the prompt:

I want to see the most impressive images of an upside down reality. From objects, to landscapes, or anything in between, what kinds of inceptions can you create.

[-] j4k3@lemmy.world 2 points 6 months ago

::: spoiler Nice job. Thanks for posting. Basic examples to guide anyone struggling with the abstractness and unconstrained artistic freedom:

https://duckduckgo.com/?q=upside+down+inception&iax=images&ia=images

By no means are you restricted to such images as these. It could mean anything you chose. You have absolute artistic freedom that is unconstrained. This is about art and creativity, not just novelty. Heck, you can even abstract my abstraction and call "Down Under" upside down and go Aussie. My point is that there are absolutely no limits, just that the theme must convey "The Up Side Down" in whatever way you imagine.

This is also an opportunity to show off skills with LoRA's and such. In the last challenge several entries showed a desire to explore the edges of complexity and possibilities. I hope this challenge pushes those buttons and scratches that itch.

[-] j4k3@lemmy.world 2 points 6 months ago* (last edited 6 months ago)
  • Cloudless
  • Vote: 7
  • Most upvoted: +3 points
  • Prompt: +1 point
  • Total: 11 points
  • Tied for first. Thanks for participating! (1/8)

Edit: big dummy forgot one point. The heck if I know how 1st ties get handled. Deathrace is my rule. first to post a #38 wins the opportunity - my rule... on the fly.

[-] PeriodicallyPedantic@lemmy.ca 6 points 6 months ago* (last edited 6 months ago)

It took me a long time to actually come up with a decent submission. I wasn't super clear what the theme is, and then once I came up with ideas I had a hard time actually generating what I wanted lol.

Here I was going for both the literal "upside down" aspect, and also Stranger Things "upside down" vibes.

I was really hoping that I could get the dude in the reflection to be removed or be different, but I couldn't seem to get both that and everything else I wanted in the same image

MS copilot designer prompta man in a hazmat suit falls through a puddle into a spooky alternate reality. in contrast, the reflection in the puddle shows a cheerful metropolis. the man is still falling through the air. the scene is very surreal.

[-] PeriodicallyPedantic@lemmy.ca 4 points 6 months ago* (last edited 6 months ago)

here are some other ideas I had

City built on the interior surface of a hollow earth. It would just NOT make the city upside down without anything below it but a glowing ball.

And a bit more figurative interpretation of theme: a world in where vampires donate blood instead of suck blood, but I couldn't ever make it clear that's what the vampire was doing

Also a stepford suburb where the townspeople were super serious about having you drink your milk. Idky but this was the first thing I thought of, but in the end I thought this was too far from the prompt

[-] cloudless@lemmy.cafe 2 points 6 months ago

I like the vampire one. Looks like he is trying to steal blood from the patient who is unconscious.

[-] PeriodicallyPedantic@lemmy.ca 2 points 6 months ago* (last edited 6 months ago)

Thanks!

He was supposed to be giving blood to the patient, but you're right that's what it looks like (which is why I didn't use it 😭)

[-] j4k3@lemmy.world 2 points 6 months ago

:::spoiler Nice one. I found this difficult too, but also very interestingly educational about what and how a model understands what it is trained on. It is one thing for a model to know what something looks like, but another thing entirely to understand it in altered context. Some models are better or worse at this kind of comprehension. Part of why I made this challenge was to see if I'm really missing any major models that are much more capable than any I have used, or perhaps other techniques I have not explored. Anyways, thanks for posting! Great image!

[-] PeriodicallyPedantic@lemmy.ca 2 points 6 months ago

Yeah I'm not advanced enough with this lol. I just use either copilot, or wombo dream (it can make some nice fantastical landscapes, but it's terrible at people). My familiarity with the different tools is really poor, I mean heavily on just being creative with my prompts and giving up on ideas that don't work out lol

[-] j4k3@lemmy.world 2 points 6 months ago* (last edited 6 months ago)
  • PeriodicallyPedantic
  • Vote:6
  • Second most upvoted: +2 pointS
  • Most unconformable: +1 point
  • OP’s favorite: +1 point
  • Prompt included: +1 point
  • Total: 11
  • First - Winner - You tied for first. Thanks for participating! (1/8)

Edit Jake forgot to add prompt to 2nd place. The heck if I know how 1st ties get handled. Deathrace is my rule. first to post a #38 wins the opportunity - my rule... on the fly.

[-] PeriodicallyPedantic@lemmy.ca 2 points 6 months ago

Thanks 😊 Let's see who wins the race lol

[-] thelsim@sh.itjust.works 4 points 6 months ago


Prompt: Versus, close-up view of an astronaut, with a large insect-like alien in the background, hanging upside down on a ceiling, in a derelict space ship corridor, dark atmosphere, low angle --ar 4:3 --stylize 70

Upside down, inverted or other kinds of viewpoints are really hard to make! Anyway, here's my attempt, it barely fits within the theme but I hope you'll like it anyway :)

[-] thelsim@sh.itjust.works 2 points 6 months ago

Some others I made along the way:

[-] j4k3@lemmy.world 2 points 6 months ago
  • Theisim
  • Vote: 4
  • Third most upvoted: +1 point
  • Prompt included: +1 point
  • Total: 6
  • Third place. Thanks for participating! I did like it. Sorry it was so hard. (3/8)
[-] thelsim@sh.itjust.works 2 points 6 months ago

That's OK, it was a challenge in the truest sense of the word :)

[-] merde@sh.itjust.works 4 points 6 months ago
[-] merde@sh.itjust.works 2 points 6 months ago* (last edited 6 months ago)
[-] j4k3@lemmy.world 1 points 6 months ago
  • merde alors
  • Vote: 4
  • Total: 4
  • Fifth place. Thanks for participating! (5/8)
[-] M0oP0o@mander.xyz 4 points 6 months ago

Prompt in the bingilator:

"Sailor moon as a planet inverted and seen from the inside of the planet looking out." Six tries.

[-] j4k3@lemmy.world 2 points 6 months ago
  • MOoPOo
  • Vote: 4
  • Prompt included: +1 point
  • Total: 5
  • Tied for Fourth place with 3 others but the first to post. Thanks for participating! (4/8)
[-] Halcyon@discuss.tchncs.de 3 points 6 months ago

M.C. Escher's Shopping Mall

It's really hard to get something upside down from the AI generator. I went for Escher's stairs and wanted to place them inside a shopping mall. As a workaround I could get the stairs inverted and added a polished floor as a mirror. Still not very convinced by the results, here's the best one I could get after several iterations.

Prompt: a scenic view inside a shopping mall with a black shiny floor that's polished like a mirror. There are several inverted stairs in a confusing upside-down architecture, similar to an image by M.C. Escher. Some people walk upside down on the underside of the stairs.

[-] j4k3@lemmy.world 2 points 6 months ago
  • Halcyon
  • Vote: 3
  • Prompt included: +1 point
  • Last two entries: +1 point
  • Total: 5
  • Tied for Fourth place. Thanks for participating! (4/8)
[-] 3volver@lemmy.world 3 points 6 months ago
[-] j4k3@lemmy.world 1 points 6 months ago
  • 3volver
  • Vote: 3
  • Total: 3
  • Sixth place. Thanks for participating! (6/8)
[-] MonkderDritte@feddit.de 3 points 6 months ago* (last edited 6 months ago)

Btw, from what film is this image?

[-] j4k3@lemmy.world 3 points 6 months ago
[-] MonkderDritte@feddit.de 3 points 6 months ago* (last edited 6 months ago)

Ah, some VR game. Thanks!
But there is a film with a similiar scene, what was it again?

[-] QuadratureSurfer@lemmy.world 2 points 6 months ago* (last edited 6 months ago)

~~Interstellar~~ "Inception", dream training scene: https://www.youtube.com/watch?v=0b-H8oQUs1A

Edit: Freudian slip.

[-] MonkderDritte@feddit.de 2 points 6 months ago
[-] Halcyon@discuss.tchncs.de 2 points 6 months ago

Elysium also had a similar scene:

[-] elvith@discuss.tchncs.de 3 points 6 months ago* (last edited 6 months ago)

A mirage from the past that never was.

I just finished the second audio book from The Sandman today and the last story in there was "Ramadan", which inspired me to create this entry.

Prompt and workflow


This image was created in two steps with SDXL in Fooocus. First step was a 21:11 (1344×704) image with the following prompt to get the mirage in the sky. I then flipped the image upside down and used outpaint on the bottom with a modified prompt


json with all parameters of the first pass


    {
    "prompt": "skyline of an historic fictive arabic city, desert palace, reminiscent of tales like One Thousand and One Nights and Aladdin, golden and colorful roofs, white houses, desert city, daylight, sun in the (blue sky:1.2)",
    "negative_prompt": "",
    "prompt_expansion": "skyline of an historic fictive arabic city, desert palace, reminiscent of tales like One Thousand and One Nights and Aladdin, golden and colorful roofs, white houses, desert city, daylight, sun in the (blue sky:1.2), intricate, elegant, highly detailed, extremely color rich, cinematic light, sharp focus, mystical, beautiful, symmetry",
    "styles": "['Fooocus V2', 'Fooocus Enhance', 'Fooocus Sharp', 'Fooocus Photograph']",
    "performance": "Quality",
    "resolution": "(1344, 704)",
    "guidance_scale": 4,
    "sharpness": 2,
    "adm_guidance": "(1.5, 0.8, 0.3)",
    "base_model": "juggernautXL_v8Rundiffusion.safetensors",
    "refiner_model": "None",
    "refiner_switch": 0.5,
    "clip_skip": 2,
    "sampler": "dpmpp_2m_sde_gpu",
    "scheduler": "karras",
    "vae": "Default (model)",
    "seed": "221192344220076966",
    "lora_combined_1": "sd_xl_offset_example-lora_1.0.safetensors : 0.1",
    "metadata_scheme": false,
    "version": "Fooocus v2.4.1"
}

json with all parameters of the second pass


{
    "prompt": "skyline of (decayed:1.2) arabic ruins, broken roofs with holes, broken walls, white dirty and sandy houses, desert city ruins, blue hour, hills in the background",
    "negative_prompt": "",
    "prompt_expansion": "skyline of (decayed:1.2) arabic ruins, broken roofs with holes, broken walls, white dirty and sandy houses, desert city ruins, blue hour, hills in the background, sharp focus, highly detailed, cinematic, elegant, intricate, professional, bright colors, surreal, dramatic, attractive, unique, cute, perfect, smart, lucky",
    "styles": "['Fooocus V2', 'Fooocus Enhance', 'Fooocus Sharp', 'Fooocus Photograph']",
    "performance": "Quality",
    "resolution": "(1216, 832)",
    "guidance_scale": 4,
    "sharpness": 2,
    "adm_guidance": "(1.5, 0.8, 0.3)",
    "base_model": "juggernautXL_v8Rundiffusion.safetensors",
    "refiner_model": "None",
    "refiner_switch": 0.8,
    "clip_skip": 2,
    "sampler": "dpmpp_2m_sde_gpu",
    "scheduler": "karras",
    "vae": "Default (model)",
    "seed": "8958039241467868771",
    "lora_combined_1": "sd_xl_offset_example-lora_1.0.safetensors : 0.1",
    "metadata_scheme": false,
    "version": "Fooocus v2.4.1"
}

[-] j4k3@lemmy.world 1 points 6 months ago
  • elvith
  • Vote: 3
  • Prompt included: +1 point
  • Last two entries: +1 point
  • Total: 5
  • Tied for Fourth place. Thanks for participating! (4/8)
[-] j4k3@lemmy.world 2 points 6 months ago

ComfyUI workflow should be embedded in the image here: https://files.catbox.moe/70w6ny.png

  • Model: incursiosMemeDiffusion_v16PDXL
  • Lora1: AltMoonPonyXL
  • Lora2: Backflip_XLPD
  • Prompt: score_9, score_8_up, score_7_up, source_anime, 1girl, upside-down, backflip, alternative sailor moon, from side, middle finger, smirk, looking at viewer, outdoors, forest, cliff, in air, adult manga version, long hair, iconic sailor outfit, short skirt, crop top, long hair
  • Neg: "", CLIP stop at layer: -2, Seed: 885720979832403, Sampler: DMPP 2M SDE Karras, Steps: 50, Cfg: 6
this post was submitted on 31 May 2024
55 points (92.3% liked)

AI Generated Images

7236 readers
260 users here now

Community for AI image generation. Any models are allowed. Creativity is valuable! It is recommended to post the model used for reference, but not a rule.

No explicit violence, gore, or nudity.

This is not a NSFW community although exceptions are sometimes made. Any NSFW posts must be marked as NSFW and may be removed at any moderator's discretion. Any suggestive imagery may be removed at any time.

Refer to https://lemmynsfw.com/ for any NSFW imagery.

No misconduct: Harassment, Abuse or assault, Bullying, Illegal activity, Discrimination, Racism, Trolling, Bigotry.

AI Generated Videos are allowed under the same rules. Photosensitivity warning required for any flashing videos.

To embed images type:

“![](put image url in here)”

Follow all sh.itjust.works rules.


Community Challenge Past Entries

Related communities:

founded 2 years ago
MODERATORS