Tsingularity/dift

[NeurIPS'23] Emergent Correspondence from Image Diffusion

41
/ 100
Emerging

This project helps visual content creators and researchers identify semantically similar points between different images, even if the objects are of different types or in varying poses. You provide two images, click a point on one, and the tool highlights the corresponding point on the other image, along with a heatmap showing similarity. This is ideal for anyone working with visual assets who needs to understand relationships between elements across diverse images.

754 stars. No commits in the last 6 months.

Use this if you need to find corresponding features between two visually different images to propagate edits or analyze relationships, like comparing a cat to a lion.

Not ideal if you're looking for exact, pixel-level matches between highly similar images for tasks like image stitching or duplicate detection.

image analysis visual content creation computer vision research visual asset management object correspondence
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 15 / 25

How are scores calculated?

Stars

754

Forks

46

Language

Python

License

MIT

Last pushed

May 14, 2024

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/Tsingularity/dift"

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