the model yolov8sbest.pt is trained using roboflow and colab and used here for inference.
the streamlit application is launched on the local server and shows available cameras for surviellance.
activate virtual environment with the said dependencies and then run the webapp by streamlit run app.py in the main directory.
each camera calls the process method from the base file and will save images with confidence>0.6 and 5 continuous frames in user directory.
