# taki0112/UGATIT

Official Tensorflow implementation of U-GAT-IT: Unsupervised Generative Attentional Networks with Adaptive Layer-Instance Normalization for Image-to-Image Translation (ICLR 2020)

Repository: https://github.com/taki0112/UGATIT
Canonical: https://ross.abutalabs.com/products/ugatit
Language: Python
License: MIT
License Family: permissive
Last push: 2021-05-20T03:23:05+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": 2596, "days_push": 1931, "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 6116, forks 1012 (observed 2026-08-28T04:09:36.088811+00:00)

## What it is
Official TensorFlow implementation of U-GAT-IT, an unsupervised image-to-image translation model using attention modules and adaptive layer-instance normalization (ICLR 2020). It includes training scripts, pretrained selfie2anime checkpoints, and dataset tooling.

## Use cases
- convert selfies into anime-style portraits
- train an unsupervised image-to-image translation model
- experiment with attention-based GANs
- reproduce U-GAT-IT paper results
- translate images between two visual domains without paired data

## When to choose
- you need unsupervised domain-to-domain image translation with large shape changes
- you want to reproduce or build on the U-GAT-IT paper
- you want the selfie2anime effect with provided checkpoints

## When to avoid
- you need modern maintained tooling - it requires TensorFlow 1.14 and Python 3.6
- you need paired image translation or production-ready inference services
- you prefer PyTorch - use the official PyTorch implementation instead

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, image-processing, computer-vision
- domain: deep-learning, computer-vision, image-processing, artificial-intelligence
- platform: python
- tags: gan, image-to-image-translation, tensorflow, attention, selfie2anime, research-code, iclr-2020, gpu

## Member repositories
- taki0112/UGATIT (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:36.088811+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 2026-08-29T17:48:00.809743+00:00, confidence not recorded.
  - readme: https://github.com/taki0112/UGATIT (fetched 2026-08-28T04:09:36.088811+00:00, sha ea245a001307)
- Data as of 2026-08-30T08:39:29.467469+00:00.