|
--- |
|
title: CircuitSketchRecognition |
|
emoji: π |
|
colorFrom: blue |
|
colorTo: blue |
|
sdk: streamlit |
|
sdk_version: 1.33.0 |
|
app_file: π€_Hello.py |
|
pinned: false |
|
license: mit |
|
--- |
|
|
|
# Circuit Sketch Recognition π |
|
|
|
## About the App |
|
The **Circuit Sketch Recognition** app demonstrates the power of AI to recognize hand-drawn circuit diagrams. By leveraging advanced models like TrOCR for text recognition and YOLOv8 for component detection, this app showcases how easily computers can understand sketches (Fine tuned on a CGHD-2304 dataset). |
|
|
|
<p align="center"> |
|
<img src="media/capture.gif" alt="Capture GIF" width="45%"/> |
|
<img src="media/upload.gif" alt="Upload GIF" width="45%"/> |
|
</p> |
|
|
|
|
|
## Features |
|
- **Upload or Capture**: Upload a picture of your circuit sketch or use your camera to capture one in real-time. |
|
- **Dual Display**: View both capture and upload options in a side-by-side format for easy access. |
|
- **Example Gallery**: Explore a gallery of recognized sketches to see the accuracy and capabilities of our AI models. |
|
|
|
## Deployment |
|
|
|
Deployed on Streamlit and Hugging Face Spaces: |
|
|
|
<a href="https://circuitsketchrecognition.streamlit.app/"> |
|
<img src="https://streamlit.io/images/brand/streamlit-mark-color.png" alt="Streamlit" width="30"/> Streamlit App |
|
</a> |
|
|
|
<a href="https://huggingface.co/spaces/edesaras/CircuitSketchRecognition"> |
|
<img src="https://huggingface.co/datasets/huggingface/brand-assets/resolve/main/hf-logo.png" alt="Huggingface" width="30"/> Huggingface App |
|
</a> |
|
|
|
## Getting Started |
|
To get started with the Circuit Sketch Recognition app, simply clone the repository and run the `π€_Hello.py` file with Streamlit: |
|
|
|
```bash |
|
git clone https://github.com/edesaras/CircuitSketchRecognition.git |
|
streamlit run π€_Hello.py |