Resist.

Inspiration

Resistors Always Fascinated Me, They are like Black-Boxes (Not Really) that are able to change the voltage of power sent through them, as someone who builds alot of hardware projects, i always have a few resistors on hand. So when i had accidentally bought a bunch of unmarked resistors, that came in a box, i knew i what i wanted to do.

What it does

You take a photo of your resistor using your camera laptop or mobile, and then upload the photo, the Deep-Learning model then will predict the resistance of the resistor in the photo with an 97% Accuracy rate, the model was built in house from an online Resistance dataset on roboflow

How we built it

Using Google Colab, We ran the python notebook, Made an Yolov8n ONN Deep learning model and hosted on an serverless architecture using AWS Lambda and AWS API Gateway

Challenges we ran into

The Project First started out as a OpenCV Model which had a measly 37% Accuracy, let alone the resistor getting detected. Now Using the Deep Learning ONN Model it Boasts a 97% Accuracy rate

Accomplishments that we're proud of

UI Revamp : The Minamilistic UI
Server Less Hosting : Using Github Actions, AWS Lambda and AWS API-Gateway to make it truly seamless from the moment i push code to the time its deployed
Switching from OpenCV to an YoLov8n Based ONN Model