A newer version of the Streamlit SDK is available:
1.40.1
metadata
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).
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:
Streamlit App
Huggingface App
Getting Started
To get started with the Circuit Sketch Recognition app, simply clone the repository and run the π€_Hello.py file with Streamlit:
git clone https://github.com/edesaras/CircuitSketchRecognition.git
streamlit run π€_Hello.py