Shape detector

Inspiration

I wanted to use web development tools and I decided to try built a web site for the first time during this hackathon and to see how to integrate neural network with in a web site looked interesting to us.

What it does

Detects what shape is in the image between a circle,square,triangle and star using neural network.

How we built it

We wrote our project in Visual Studio Code,in HTML,CSS and JavaScript,used Git and Github to manage our joined work. Wrote an image classification in python utilizing Keras library ,than converted the model to TensorFlow.js jason file. Used Node.js and Express to build a server and uploaded our web site to Heroku.

Challenges we ran into

Building the page layout was challenging,how to use HTML and CSS and adding JavaScript functions. The neural network was a challenge to improve accuracy.

What we learned:

We learned how to upload a website to Heroku,built web page with CSS and HTML,Node.js and JavaScript.
how to convert a model that was built with Keras to TensorFlow.js.

What next to Shape Detector

We'll improve the model accuracy and upgrade our web site.

Built With

• css
• express.js
• git
• github
• heroku
• html
• javascript
• node.js
• python
• shell