Repository Details
Shared by
HelloGitHub Rating
10.0
1 ratings
AI Tool for Screen Parsing
Past 7 days Received 16 stars ✨
Use arrow keys to explore dates, Home and End for the first and last date, Escape to dismiss.| Date | Star | Daily change |
|---|---|---|
| 周二 | 25,613 | — |
| 周三 | 25,614 | +1 |
| 周四 | 25,622 | +8 |
| 周五 | 25,622 | 0 |
| 周六 | 25,622 | 0 |
| 周日 | 25,625 | +3 |
| 周一 | 25,629 | +4 |
25.6k
Stars
No
Chinese
Jupyter Notebook
Language
Yes
Active
16
Contributors
233
Issues
Yes
Organization
None
Latest
2k
Forks
CC-BY-4.0
License

This is an open-source screen parsing tool by Microsoft that can convert UI screenshots into structured and easily manageable elements. Developed in Python, it leverages models such as YOLO, BLIP2, and Florence to achieve accurate icon recognition and generate descriptive text. It also supports integration with mainstream large language models (e.g., GPT-4V) and is suitable for developing desktop automation applications.
Comments
Rating:
No comments yet