Military vehciles detection

Inspiration

Wanted to do a deep learning project could be useful to military

What it does

It tells whether there is a military tank/truck/artilary in an image

How we built it

We created our own dataset of tanks,artilary and trucks. Added car dataset to it.We then built a neural network architecture using our modified version of Lenet-5 architecture using keras and tensorflow

Challenges we ran into

• We had to figure out how to batch download files from google
• Create our own dataset and clean it and label it
• Select a neural network architecture and modify it for our purpose ## Accomplishments that we're proud of Test Accuracy = 99% ## What we learned
• How to create a dataset and label it.
• Learned generators in keras and how to make a model of your own.
• Learned YOLO, RCNN object recognition algorithms ## What's next for Military vehciles detection
• If given a dataset of drone surveillance we can use it to predict where military establishments are.

Built With

• keras
• python
• tensorflow