# google/style-aligned

Official code for "Style Aligned Image Generation via Shared Attention"

Repository: https://github.com/google/style-aligned
Canonical: https://ross.abutalabs.com/products/style-aligned
Language: Python
License: Apache-2.0
License Family: permissive
Archived: true
Last push: 2023-12-29T21:26:34+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 72
- inputs: {"age_days": 1017, "days_push": 978, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, archived
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1315, forks 95 (observed 2026-08-28T04:04:20.585127+00:00)

## What it is
Official research code for 'Style Aligned Image Generation via Shared Attention', implementing style-consistent image generation with diffusion models like SDXL via shared attention mechanisms. It provides example notebooks and demos for style-aligned generation, style transfer from reference images, ControlNet conditioning, and panorama generation.

## Use cases
- generate a batch of images with a consistent visual style
- transfer the style of a reference image to new generated images
- create style-aligned images conditioned on depth maps with ControlNet
- generate style-consistent panoramas with MultiDiffusion
- reproduce results from the Style Aligned paper

## When to choose
- you want style-consistent outputs from diffusion models without training or fine-tuning
- you already use Hugging Face Diffusers and want to add shared-attention style control
- you want reference implementations of the Style Aligned paper's method

## When to avoid
- you need a production-supported product - it is explicitly not an officially supported Google product
- you need features still on the TODO list like multi-style MultiDiffusion or DreamBooth integration
- you don't work with diffusion models or lack a GPU

## Facets
- artifact type: library
- maturity: experimental
- function: image-processing, machine-learning, deep-learning
- domain: artificial-intelligence, image-processing, deep-learning
- platform: python
- tags: diffusion-models, stable-diffusion, sdxl, shared-attention, style-transfer, text-to-image, research-code, diffusers, gpu

## Member repositories
- google/style-aligned (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:20.585127+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:48:41.134615+00:00, confidence not recorded.
  - readme: https://github.com/google/style-aligned (fetched 2026-08-28T04:04:20.585127+00:00, sha 101ffea4c07a)
- Data as of 2026-08-30T08:39:29.467469+00:00.