taco-group

taco-group / PISCO

Public

PISCO: Precise Video Instance Insertion with Sparse Control

xiangbogaobarry.github.ioPISCO agent filming image-to-video llm text-to-video
47
1
100% credibility
Found Feb 17, 2026 at 37 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

PISCO enables precise insertion of objects into existing videos using sparse frame masks and optional depth maps for realistic motion and integration.

How It Works

1
🔍 Spot amazing video edits

You see stunning videos online where toys or characters appear perfectly in real footage at exact moments, sparking your curiosity.

2
📥 Grab the video editor

Download this free tool that makes precise object insertion into videos super easy, no skills needed.

3
🎥 Load your video

Pick your favorite video clip and the image of the object you want to add, like a cute toy.

4
🎯 Mark special spots

Simply paint quick masks on just a few key frames where you want the object to show up, and add depth info if you have it.

5
✨ Hit generate magic

Watch as the tool smartly inserts your object smoothly across the whole video, matching motion and lighting perfectly.

✅ Enjoy your creation

Play back your customized video with seamless additions, ready to share with friends and family.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 37 to 47 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is PISCO?

PISCO lets you precisely insert objects or instances into existing videos using sparse control signals, like keypoints or trajectories, instead of dense masks. Built in Python on diffusion models, it handles tasks such as adding items to scenes, repositioning elements, resizing, or simulating interactions, with ready-to-run inference for 480p/720p outputs via simple scripts. Developers get Hugging Face models (1.3B and 14B params) and demos showing seamless video edits, perfect for precise control in video generation.

Why is it gaining traction?

It stands out with sparse pisco control for instance insertion, delivering realistic results without heavy annotations—think dropping a pisco sour bottle into a bar scene via Peru-inspired sparse cues. Compared to full-video diffusion tools, PISCO's precision on motion and occlusion makes edits feel natural, and one-command inference (after pip install) hooks users experimenting with video control. Early adopters praise the pisco sour rezept-like simplicity for creative tweaks.

Who should use this?

Video AI researchers tuning sparse control for diffusion pipelines, ML engineers prototyping object insertion in apps like AR ads or e-commerce visuals (pisco kaufen demos at Edeka/Rewe), and indie game devs needing dynamic scene edits. Ideal for those handling pisco Peru authenticity in motion videos without rebuilding from scratch.

Verdict

Grab it if you're into precise video insertion—inference works out of the box with solid demos—but with 34 stars and 1.0% credibility score, it's early-stage; expect tweaks for production. Strong start for Python diffusion tinkerers.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.