Waste Predictor

Web App HomePage
Image Classification

Inspiration

What it does

The User uploads an image in the webapp and ultimately it can classify it in two categories 'compostable' and 'Recyclable' by finding the accuracy from the model that we have created.

How we built it

We first made the model by using Auto ML vision Image Classifier supported by google cloud. Then we set the confidence threshold appropriately to give maximum accuracy. At last, we wrapped our model in the desired UI and then deployed the result in our Web App using the API tensorflow.js.

Challenges we ran into

The major challenge we faced was the huge dataset. Extracting the images as a matrix increases the features in the dataset exponentially. So training the model took a lot of time.

Accomplishments that we're proud of

After tackling all the difficulties, we have successfully completed our first hackathon and also used the google cloud supported algorithm for the first time.

What we learned

Deploying a Machine learning model in the web application so any user can easily connect with it and use the result of the model without running the model.

What's next for Waste Predictor

We can use the result of the model and make the prediction for real time. So a general use of this model can be that the model being deployed in the camera and it is guarded near the dustbin so any user came to dump garbage can easily get to know in which of the dustbin he has to throw whether it should be recyclable or compostable.