mingyuliutw/UNIT

Unsupervised Image-to-Image Translation

50
/ 100
Established

This project transforms an image from one visual style or domain into another, without needing paired examples for training. For instance, you can take a daytime street scene and generate a nighttime version, or change a dog's breed in a photo. It takes a collection of images from a source domain and a collection from a target domain, and outputs translated images. This tool is useful for artists, designers, or visual content creators who need to generate diverse image variations or augment datasets.

2,030 stars. No commits in the last 6 months.

Use this if you need to translate the visual style or characteristics of images from one domain to another (e.g., snowy to summery, day to night, or changing animal breeds) and don't have exact 'before and after' image pairs for training.

Not ideal if you need a more performant implementation or multimodal (many-to-many) image translation, in which case the MUNIT or Imaginaire projects might be better suited.

image-style-transfer visual-effects content-creation digital-art dataset-augmentation
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 24 / 25

How are scores calculated?

Stars

2,030

Forks

363

Language

Python

License

Last pushed

Sep 02, 2021

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/mingyuliutw/UNIT"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.