Sign In

UFOMaker

2.4k

Download

1 variant available

SafeTensor

36.13 MB

Verified:

Type
LyCORIS
Stats

497

393

48

Reviews
Published

Mar 1, 2024

Base Model

SD 1.5

Hash
AutoV2
253D359987
Trigger Words
ufomaker
flying saucer
spaceship
spacecraft
ufo
exhaust fumes
1950s (style)
default creator card background decoration
Uploads - 109

109

Reactions - 11721

11.7K

Generations - 225565

225.6K

Maintenance Mode Badge May 2023
00000-123654.png

UFOMaker is a tool for generating all types of UFOs and spaceships. It was trained on an in-house developed dataset of 360 designs with interesting concept features (this is my most time-consuming model project yet). The dataset was partially created as Blender ControlNet inputs for AI generation or hand retouched from generated sketches:

The model either fixes the original input or makes it more interesting.

Use Hires. fix (4x_RealisticRescaler, 4x-Ultrasharp upscalers with denoising strength around 0.3-0.6) for best details in txt2img!

Experiment with weights around 0.1-1 (see above image, the lower values will modify the design, higher will create more asset type picture) , CLIP SKIP 1-2 . The model works with ufo and spaceship tokens (try to use a combination ufo|spaceship, ufo:spaceship:ufo, and so on). It will also style a combined vehicle.

LyCORIS/LoHA experiment trained 512x512, NR 64 + NA 32 dimensions. I am not aiming for technical realism with this series, the models are more focused on design ideas. If you want to learn more about AI-augmented design and algorithmic art, check my web pages.

For my notes on training, research, and use of stable diffusion models, check https://sandner.art/articles/ai/

If you can afford it, support my work by buying me a coffee to keep those costly render cores go and to keep creative flow of this content: https://ko-fi.com/airesearch

If you can not, review an rate the model to help others to find it.

Plans for further development

  • Train for higher res with updated dataset

  • Actions and technical details training (for next versions)

  • Optimize the training

  • Optimize the dataset

  • Create more focused Light version

Have fun rendering!