Gender Rover

EMBEDDED DEVELOPMENT

ECHIPA : FizAr

Inspiration

What brought me and my teammate to do this project was the fact we wanted to learn how face recognition works but without any fancy code or methods that we don't understand. So, we thought the easiest way to get around this problem is by making the raspberry pi to choose between who is a man or a woman. If it is a man the rover goes forward for 12 seconds and if after that the next person or image shown to him is a woman it will print a short message (FEMALE).

What it does ?

Well it does only the face recognition part between man and woman. In the future it can be perfected even in a smart way to unlock your car, who knows :)) .

How I built it

Instead of the Pi Camera I have used my own Microsoft usb camera with OpenCV. Under the chasis are four motors tied to a single motor driver (L298N). For processing the facial recognition part I have used Intel Movidius Neural Stick.

Challenges I ran into

My first challenge that I ran into was my lack of experience in neural networks. But after two days without sleep I have achieved more experience in it. As for the second challenge that I ran into was the idea part but I sticked to something simple that I can understand and improve it in the future.

Accomplishments that I'm proud of

I am proud of using the Raspberry Pi and learning how to use neural networks.

What I learned

• Neural networks
• How to think the same as a machine does

What's next for Gender Rover

For this rover I am planning to expand it to some sort of lock for cars with facial recognition but till then I am focusing on learning more about neural networks.

Built With

• python