ML solution to detect fight that occur using camera streaming.

• IMSOO Git Repo
• YOLOv8 Pose Estimation

Install ananconda environment

Create new environment with python 3.10.12

conda create -n fdet python=3.10.12 pip

Enter the environment

conda activate fdet

Clone the repo, enter the directory

git clone https://github.com/erwinyo/Fight-Detection-Yolov8-Pose

cd Fight-Detection-Yolov8-Pose

Install the requirements.txt file

pip install -r requirements.txt

Run the main script

python3 main.py

Go to http://127.0.0.1:8000

if the text "This is a testing page. It tells you this is working :)" showing then the server running.

Go to http://127.0.0.1:8000/start?video_input=https://sample-videos.com/video321/mp4/720/big_buck_bunny_720p_20mb.mp4

If you see the result. Then it running

You edit the input to be RTSP connection with your CCTV

If you want to customize this, you can modified the following environment variables to your .env file