# jamriska/ebsynth

Fast Example-based Image Synthesis and Style Transfer

Repository: https://github.com/jamriska/ebsynth
Canonical: https://ross.abutalabs.com/products/ebsynth
Homepage: https://ebsynth.com
Language: C
License Family: other
Topics: image-synthesis, style-transfer, image-analogy, texture-synthesis
Last push: 2023-06-14T21:31:49+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 3062, "days_push": 1176, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1674, forks 206 (observed 2026-08-28T04:05:20.283004+00:00)

## What it is
Ebsynth is a fast example-based image synthesis tool that performs style transfer, guided texture synthesis, inpainting, and super-resolution using non-parametric patch-based algorithms rather than neural networks. It is available as a command-line tool with CPU and CUDA backends, and also powers a commercial VFX web/desktop app for propagating painted keyframes across video.

## Use cases
- transfer artistic style from one image to another
- synthesize textures guided by segmentation maps (texture-by-numbers)
- stylize 3D render passes like an artist would (StyLit workflow)
- propagate a painted keyframe across video frames for rotoscoping and animation
- content-aware inpainting of images
- example-based super-resolution preserving fine detail

## When to choose
- you need crisp, detail-preserving synthesis without neural networks or GPU training
- you want to stylize video by painting only keyframes
- you need guided texture synthesis from a source photo plus segmentation
- you want a lightweight CLI with optional CUDA acceleration

## When to avoid
- you need generative AI models trained on large datasets
- you need a GUI in the open-source tool itself (the CLI is text-only)
- your project requires a permissive license — the repo has no license file
- you need active feature development (releases are infrequent)

## Facets
- artifact type: cli-tool
- maturity: maintenance
- function: image-processing, graphics
- domain: image-processing, graphics, computer-vision
- platform: windows, cli, cpp
- tags: style-transfer, texture-synthesis, image-analogy, patch-based-synthesis, non-photorealistic-rendering, vfx, cuda, linux, macos

## Member repositories
- jamriska/ebsynth (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:20.283004+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 2026-08-30T03:42:32.157378+00:00, confidence not recorded.
  - readme: https://github.com/jamriska/ebsynth (fetched 2026-08-28T04:05:20.283004+00:00, sha 32036ee06a7d)
  - homepage: https://ebsynth.com (fetched 2026-08-29T11:15:20.530706+00:00, sha a528a34197d4)
- Data as of 2026-08-30T08:39:29.467469+00:00.