AI Glossary

Style transfer

Also called neural style transfer

What does “Style transfer” mean?

Style transfer: Re-rendering one image in the visual style of another while keeping its content. Early methods used convolutional neural networks to separate "content" from "style" features; today the same effect is often achieved with image-to-image diffusion, reference-image inputs or style LoRAs.

Style transfer visual guide: Re-rendering one image in the visual style of another while keeping its content. Early methods used convolutional neural networks to separate "content" from "style" features; today the same effect is often achieved with image-to-image diffusion, reference-image inputs or style LoRAs.
Visual guide Open full size (new tab)

What is an example of “Style transfer”?

Making a city photo look like it was painted with the brushstrokes and palette of a particular painting.

Keep learning

See also

Image-to-imageTermLoRA (image models)TermConvolutional neural networkTermText-to-imageTerm

Same topic

Temporal consistencyTermSpeech-to-textTermSeedTermText-to-speechTermSampling stepsTerm