Pixel Art Fixer is an open-source tool that uses classical image processing to convert fake, off-grid pixel art into real, grid-aligned pixel art without requiring a neural model.

Stars

238

7-day growth

No data

Forks

20

Open issues

0

License

MIT

Last updated

2026-07-15

AI repository intelligence
FR-AI / ANALYSIS

Why it is worth attention

It achieves state-of-the-art accuracy among non-neural methods by employing multiple independent detectors with consensus voting and a two-stage reconstruction that decouples structure from color.

Who it is for

  • Pixel artists and game developers needing to clean up AI-generated or upscaled pixel art
  • Developers integrating image processing into pixel-art pipelines
  • Content creators restoring old or compressed pixel art assets
  • Researchers or hobbyists interested in grid-detection algorithms

Use cases

  • Restoring true pixel art from AI-upscaled or mushy images
  • Preparing sprites for game engines that require exact grid alignment
  • Batch-fixing pixel art in a folder with the Python or Rust CLI
  • Integrating pixel art repair into a web service via the provided API

Strengths

  • Outperforms existing tools (PixelDetector, unfake.py, PixelSnapper) with 77% exact native resolution recovery and 94% grid alignment accuracy across 43 distortion categories
  • Uses three cheap detectors (autocorrelation, run-lengths, self-similarity) plus a heavier evidence stack for robust consensus
  • Handles non-integer cell sizes, sub-pixel drift, anti-aliasing, JPEG artifacts, and warped grids
  • Two-stage reconstruction produces crisp lines and accurate colors by separating structure voting from color sampling

Considerations

  • Purely image-processing based; cannot recover images with extreme damage (e.g., cells under ~3 px, heavy dithering) where a neural version is needed
  • The neural version is hosted externally on retrodiffusion.ai and not included in this repository
  • Input limited to PNG and JPEG, with a maximum resolution of 4 megapixels

README quick start

Usage

Description

Convert images into pixel art

Related repositories

Similar projects matched by category, topics, and programming language.

lopopolo
Featured
lopopolo GitHub avatar

harness-engineering

Harness Engineering is a methodology for improving coding agent outputs by carefully crafting the environment around them—providing curated context, tools, and executable constraints that encode an organization’s nonfunctional requirements and cumulative lessons.

AI & Machine LearningAI Agents
2,390
slvDev
Featured
slvDev GitHub avatar

esp32-ai

A 28.9 million parameter language model runs on an $8 ESP32-S3 microcontroller entirely on-device, generating simple stories at about 9.5 tokens per second.

AI & Machine LearningLarge Language Models
1,960
littledivy
Featured
littledivy GitHub avatar

mimic

mimic captures traffic from any iOS or web app and automatically generates a Python client library that lets you call the app's API like a regular library.

AI & Machine Learning
1,482