Motivation and Goals

• There are several gun-related crimes happening in the USA. So, I attempted to make a website using yolov10, which is a robust computer vision algorithm that can detect objects with laser precision.
• Another goal of mine to make this web-app is to learn how CI/CD pipeline, Docker, and AWS works. I deployed the project on AWS using EC2 and ECR, since I don't have a premium subscription, I had to take it down.

DEMO

• video link

Tech Stacks

• Vanila JS for frontend
• Flask for backend

Machine learning pipeline

• There are three main components

1. Data Ingestion
2. Data Validation
3. Model Trainer

Running on local machine

1. you should clone the repo using https://github.com/devPach4545/Gun-detection.git and go to the Gun-detection dir
2. Create an anaconda or a virtual environment
3. Use pip install -r requirements.txt to install the dependencies
4. In your command line, type python app.py and run the application.
5. Also, make sure you use the avaliable port and host for your flask app.
6. Show the video above to detect a gun

Note:

1. I do not have an access to a strong gpu, therefore I have trained the model for only 80 epochs, which is why you may not see a remarkable accuracy.

Results

•