Challenges we ran into
A challenge that we ran into was that the size of the dataset was not large enough, so we only have a limited amount of objects within the model, but we will be adding to that soon.
Accomplishments that we're proud of
An accomplishment that we are proud of is making the machine learning model 92% accurate with the objects that we used to train it.
What we learned
The main takeaway for us is how to use a TensorFlow model within a web application.
What's next for Vision Scope
Our future for VisionScope AI is to implement the web application into an app, and make users be able to take pictures of household items, and have it classify into Trash, Recycling, or Compost.
Built With
- html
- javascript
- machine-learning
- tensorflow
Log in or sign up for Devpost to join the conversation.