Spaces:
Build error
Build error
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,12 +1,30 @@
|
|
| 1 |
-
|
| 2 |
-
|
| 3 |
-
|
| 4 |
-
|
| 5 |
-
|
| 6 |
-
|
| 7 |
-
|
| 8 |
-
|
| 9 |
-
|
| 10 |
-
---
|
| 11 |
-
|
| 12 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# Sign Language Project
|
| 2 |
+
|
| 3 |
+
⚡ An AI-powered system for translating Russian Sign Language (RSL) gestures into fluent, grammatically correct Russian text.
|
| 4 |
+
|
| 5 |
+
## 🚀 Overview
|
| 6 |
+
|
| 7 |
+
The Sign Language Project integrates computer vision and natural language processing to create an end-to-end pipeline:
|
| 8 |
+
|
| 9 |
+
- **Gesture Recognition**: Uses ResNet-3D and OpenCV to identify continuous RSL video input.
|
| 10 |
+
- **Linguistic Correction**: Applies GPT-3.5-Turbo via API to convert gloss output into fluent Russian sentences.
|
| 11 |
+
|
| 12 |
+
This project aims to assist educators, students, and developers working with RSL by automating the translation of sign language into readable text.
|
| 13 |
+
|
| 14 |
+
## 🔬 Technical Highlights
|
| 15 |
+
|
| 16 |
+
- **ResNet-3D**: For video frame-based gesture recognition.
|
| 17 |
+
- **OpenCV**: For video preprocessing and gesture extraction.
|
| 18 |
+
- **GPT-3.5-Turbo Integration**: For morphosyntactic correction and fluent sentence generation.
|
| 19 |
+
- **Streamlit Interface**: For easy testing and demonstration.
|
| 20 |
+
|
| 21 |
+
## 👥 Community & Usage
|
| 22 |
+
|
| 23 |
+
- Publicly available on Hugging Face.
|
| 24 |
+
- Tested by developers and educators exploring sign language technology.
|
| 25 |
+
- Recognized in expert evaluations as an example of applied AI for assistive technology.
|
| 26 |
+
|
| 27 |
+
## 📝 License
|
| 28 |
+
|
| 29 |
+
This project is released under the [MIT License](LICENSE).
|
| 30 |
+
|