garibida/cross-image-attention
A zero-shot appearance transfer method using cross-image attention in stable diffusion to combine the structure of one image with the style of another.
Not currently ranked — collecting fresh signals.
star history
This is the official implementation of a SIGGRAPH 2024 research paper on cross-image appearance transfer. The method leverages self-attention layers in pre-trained text-to-image diffusion models to establish semantic correspondences between two images, transferring visual appearance without requiring any training or optimization. By combining queries from a structure image with keys and values from an appearance image during the denoising process, it generates new images that retain the structure of one input while adopting the visual style of another.
Frequently asked
- What is garibida/cross-image-attention?
- A zero-shot appearance transfer method using cross-image attention in stable diffusion to combine the structure of one image with the style of another.
- Is cross-image-attention open source?
- Yes — garibida/cross-image-attention is open source, released under the MIT license.
- What language is cross-image-attention written in?
- garibida/cross-image-attention is primarily written in Python.
- How popular is cross-image-attention?
- garibida/cross-image-attention has 404 stars on GitHub.
- Where can I find cross-image-attention?
- garibida/cross-image-attention is on GitHub at https://github.com/garibida/cross-image-attention.