Iscream Project

Try it out on Kaggle: https://www.kaggle.com/code/aadipatangi/freyhacks-image-setup

Inspiration

We were inspired to build an ice-cream classifier by the theme of summer and what we appreciate most about it, a good ol' scoop of ice cream.

What it does

Using Tensorflow, we built a model that can classify between 4 different brands of ice cream: Häagen-Dazs, Breyers, Ben & Jerry's, and Talenti.

How we built it

We built our model by creating a convolutional neural network with 3 layers - 1 Flatten and 2 dense layers. We trained it based on a dataset of ice cream images from Kaggle [https://www.kaggle.com/datasets/tysonpo/ice-cream-dataset]

Challenges we ran into

Some challenges that we ran into were installing packages. For a reason that we are still unclear about, Tensorflow was not installing on our Visual Studio Code which is why we switched to Kaggle. Kaggle was sluggish and sometimes it was frustrating to wait for it to catch up.

Accomplishments that we're proud of

We are proud that we created a functional neural network. We are both newbies in the field of machine learning.

What we learned

We learned how to build a neural network. We also learned how to properly do time management. I personally learned how to make a youtube video.

What's next for Iscream or NotCream

We want to make it available on the web possibly using flask.