crowsonkb / k-diffusion

Karras et al. (2022) diffusion models for PyTorch
MIT License
2.26k stars 372 forks source link

[Feature request] Combination of Inpainting and ImgToImg Generation for finetunig #29

Open Serasul1984 opened 1 year ago

Serasul1984 commented 1 year ago

Combination of InPainting and ImgToImg Generation for fine tuning:

You would go to InPainting select an option to upload an second layer you can cut,trim,rotate,flip,make transparent with an strength slider and scale the second layer than you place the second layer where you want over the first layer(picture u use for imageToimage generation. So you can add good generations of details like (faces,hands,eyes,armor patterns and so on) so small parts of an picture to the picture you want to use for generating an totally new one. Nor with the mask option you also only mask the positions of the second layer so the generating would fit in the new second layer on the first without destroying the first layer you like too.

This process would save time and energy to get what you had in mind when you type the prompt.