Yolov5 Real-time Inference using Streamlit

A web interface for real-time yolo inference using streamlit. It supports CPU and GPU inference, supports both images and videos and uploading your own custom models.

Live Demo

Features

• Caches the model for faster inference on both CPU and GPU.
• Supports uploading model files (<200MB) and downloading models from URL (any size)
• Supports both images and videos.
• Supports both CPU and GPU inference.
• NEW: supports rtsp webcam stream
• Supports:
  • Custom Classes
  • Changing Confidence
  • Changing input/frame size for videos

How to run

Make shure you have installed Python and PIP. You also need:

1. Ultralytics
   • pip install ultralytics
2. Streamlit
   • pip install streamlit

After cloning the repo:

1. Install requirements
   • pip install -r requirements.txt
2. Add sample images to data/sample_images
3. Add sample video to data/sample_videos and call it sample.mp4 or change name in the code.
4. Add the model file to models/ and change cfg_model_path to its path.

git clone https://github.com/moaaztaha/Yolo-Interface-using-Streamlit
cd Yolo-Interface-using-Streamlit
streamlit run app.py

To-do Next

• Allow model upload (file / url).
• resizing video frames for faster processing.
• batch processing, processes the whole video and then show the results.

References

https://discuss.streamlit.io/t/deploy-yolov5-object-detection-on-streamlit/27675