# yangxy/PASD

[ECCV2024] Pixel-Aware Stable Diffusion for Realistic Image Super-Resolution and Personalized Stylization

Repository: https://github.com/yangxy/PASD
Canonical: https://ross.abutalabs.com/products/pasd
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2024-09-23T02:58:40+00:00

## Health v2 (maintenance only)
Score: 28/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 78
- inputs: {"age_days": 1094, "days_push": 709, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1021, forks 65 (observed 2026-08-28T04:03:15.752670+00:00)

## What it is
PASD (Pixel-Aware Stable Diffusion) is a Python research codebase implementing an ECCV 2024 method for realistic image super-resolution and personalized stylization using diffusion models. It provides pretrained weights, SDXL support, colorization, tiled latent upscaling for high-resolution images, and Gradio/Colab demos.

## Use cases
- upscale low-resolution photos to high resolution realistically
- enhance real-world degraded images with diffusion models
- stylize photos with a personalized stable diffusion model
- colorize grayscale images
- upscale ultra high-resolution images with tiled inference
- control detail level of super-resolution output

## When to choose
- you need state-of-the-art diffusion-based realistic image super-resolution
- you want pretrained PASD or PASD-SDXL weights for restoration or stylization
- you need to upscale very large images via tiled latent inference
- you want a research reference implementation from an ECCV 2024 paper

## When to avoid
- you need a lightweight non-GPU image upscaler for production
- you want a polished end-user application rather than research code
- you lack a GPU or cannot run large diffusion models
- you need fast real-time processing

## Facets
- artifact type: library
- maturity: active
- function: image-processing, stable-diffusion, machine-learning, deep-learning
- domain: image-processing, computer-vision, artificial-intelligence, deep-learning
- platform: python, cross-platform
- tags: super-resolution, stylization, diffusion-models, eccv2024, image-restoration, research-code, gpu, linux

## Member repositories
- yangxy/PASD (main) score 28

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:15.752670+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 2026-08-30T07:08:52.727376+00:00, confidence not recorded.
  - readme: https://github.com/yangxy/PASD (fetched 2026-08-28T04:03:15.752670+00:00, sha eba7eabaf39a)
- Data as of 2026-08-30T08:39:29.467469+00:00.