#244 · Primary category: Computer Vision
ghost
A new one shot face swap approach for image and video domains
Project last updated:02/25/25
GitHub Stars
1.6K
Forks
308
Contributors
5
License
Apache-2.0
Why we included this project
GHOST is a one-shot face swap model: give it a single source photo and a target image or video, and it transfers that identity across while holding onto the target's expressions and head movement. It comes out of the ai-forever research group with a paper on IEEE Xplore, and the repo includes an inference script, pretrained weights, and a Colab demo, so you can try it before wiring it into anything bigger. The training code and loss controls are also there, useful if you want to fine-tune it on your own portrait set or change how strongly the result favors source identity versus target appearance. One thing worth weighing for production use: the authors explicitly restrict GHOST to ethical, consented scenarios and research, so your pipeline needs to sit comfortably inside that boundary.
Articles for this project
No articles for this project yet.
To suggest a topic or contribute an article, contact us.
Related projects in this category
opencv
Open Source Computer Vision Library
RuView
π RuView turns commodity WiFi signals into real-time spatial intelligence, vital sign monitoring, and presence detection — all without a single pixel of video.
PaddleOCR
Turn any PDF or image document into structured data for your AI. A powerful, lightweight OCR toolkit that bridges the gap between images/PDFs and LLMs. Supports 100+ languages.
MinerU
Transforms complex documents like PDFs and Office docs into LLM-ready markdown/JSON for your Agentic workflows.
tesseract
Tesseract Open Source OCR Engine (main repository)