# sczhou/CodeFormer

[NeurIPS 2022] Towards Robust Blind Face Restoration with Codebook Lookup Transformer

Repository: https://github.com/sczhou/CodeFormer
Canonical: https://ross.abutalabs.com/products/codeformer
Language: Python
License: NOASSERTION
License Family: other
Topics: codebook, codeformer, face-enhancement, face-restoration, pytorch, super-resolution, vqgan, restoration
Last push: 2025-11-18T12:03:30+00:00

## Health v2 (maintenance only)
Score: 46/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 52, release rhythm 8, longevity 100
- inputs: {"age_days": 1534, "days_push": 288, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 18117, forks 3719 (observed 2026-08-28T04:11:26.327665+00:00)

## What it is
CodeFormer is a PyTorch-based blind face restoration model using a codebook lookup transformer, published at NeurIPS 2022. It restores and enhances degraded face images with support for super-resolution, inpainting, colorization, and video input.

## Use cases
- restore old blurry face photos
- enhance low-quality faces in images
- upscale and denoise faces in videos
- fix degraded AI-generated face art
- colorize black-and-white face images
- inpaint missing regions of aligned face crops

## When to choose
- you need state-of-the-art blind face restoration on degraded photos
- you want pretrained checkpoints plus inference and training code
- you need face enhancement for images or videos in a Python pipeline

## When to avoid
- you need general-purpose image restoration beyond faces
- you lack a GPU or want lightweight real-time processing
- you require a permissive open-source license for commercial use

## Facets
- artifact type: library
- maturity: stable
- function: image-processing, machine-learning, deep-learning
- domain: computer-vision, image-processing, deep-learning, artificial-intelligence
- platform: python, cross-platform
- tags: face-restoration, super-resolution, blind-face-restoration, vqgan, transformer, codebook, face-inpainting, colorization, pytorch, neurips-2022, gpu

## Member repositories
- sczhou/CodeFormer (main) score 46

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:11:26.327665+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 2026-08-29T17:02:09.242976+00:00, confidence not recorded.
  - readme: https://github.com/sczhou/CodeFormer (fetched 2026-08-28T04:11:26.327665+00:00, sha e395216d5a3f)
- Data as of 2026-08-30T08:39:29.467469+00:00.