Haoming02/FIRST

Frame Interpolation Refined with Stable Diffusion via Control Net

35
/ 100
Emerging

When you have video footage where objects move too quickly or the scene changes too drastically, traditional frame interpolation often struggles, leading to blurry or 'ghosting' effects. This project takes your original video frames and any preliminary interpolated frames (like those from Flowframes) and refines the missing frames. It leverages advanced generative AI to create smoother, more coherent transitions, resulting in a cleaner, more professional-looking slow-motion or high-frame-rate video. This is ideal for video editors, animators, or content creators who need to improve the fluidity of their interpolated footage.

No commits in the last 6 months.

Use this if you need to create exceptionally smooth slow-motion or high-frame-rate video from footage where there are significant differences between frames, and existing interpolation tools produce noticeable artifacts.

Not ideal if you are looking for a standalone, all-in-one video editing solution, as this tool focuses specifically on refining pre-interpolated frames.

video-editing animation content-creation post-production video-enhancement
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 6 / 25
Maturity 16 / 25
Community 13 / 25

How are scores calculated?

Stars

24

Forks

4

Language

Python

License

MIT

Category

controlnet-tools

Last pushed

Jul 05, 2023

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/Haoming02/FIRST"

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