Fr0zenCrane/Cockatiel

The official implementation of our paper "Cockatiel: Ensembling Synthetic and Human Preferenced Training for Detailed Video Caption"

20
/ 100
Experimental

Cockatiel helps generate highly detailed and human-preferred captions for videos, going beyond simple descriptions to capture fine-grained aspects like objects, camera movements, and background details. You provide a video file, and it outputs a rich, descriptive caption. This tool is ideal for researchers, content creators, or analysts who need precise, nuanced textual descriptions of video content.

No commits in the last 6 months.

Use this if you need to automatically generate detailed, human-aligned descriptions for a large collection of videos, capturing specific aspects of the video content beyond basic summaries.

Not ideal if you only need very short, generic captions or if your primary need is real-time video transcription without detailed scene analysis.

video-analysis content-creation media-research digital-asset-management AI-assisted-description
No License Stale 6m No Package No Dependents
Maintenance 2 / 25
Adoption 7 / 25
Maturity 8 / 25
Community 3 / 25

How are scores calculated?

Stars

38

Forks

1

Language

Python

License

Last pushed

May 21, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/Fr0zenCrane/Cockatiel"

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