Inspiration
After conversing with my group we quickly found out that lots of us had difficulty at times determining what is recyclable and what is not. Also, whether it was recyclable or not it is difficult to always take the best course of action at any given time.
What it does
Essentially, our application takes a picture uploaded by the user and then after the AI processes the image it provides a predictor of what the item might be. It then accordingly gives suggestions on what to do with and how to properly dispose or recycle the item.
How we built it
Initially, we started off coding the front-end and working through the application step-by-step. For example, first we built the basic CSS for the home page and divided tasks based on what our strengths were within our group. After finishing the frontend components we moved onto building our back-end. This involved importing django and training out AI models. Finally, after our AI was trained and our back-end was complete we worked on connecting it to our front-end and outputting the necessary information.
Challenges we ran into
Me and 2 of my other teammates were freshman and this was our first ever hackathon, therefore, we had barely ever touched react before and did not know what we were getting ourselves into when we went in wanting to implement AI into our project. Implementing the AI and the back-end were by far the most challenging part of the whole hackathon. We found the concepts to be extremely complex and difficult to understand.
Accomplishments that we're proud of
Me and my team are extremely proud of coming up with the idea and making it come to life. Many of us have little to no experience with front-end and back-end, therefore, the fact that we were able to have such a memorable first hackathon and a great project to go alongside with it was very fulfilling.
What we learned
My group learned how to develop using frontend and backend software. We also learned how to implement AI in the functionality of our project. Overall, we learned more than just about technical languages but also how to adapt and how to solve problems that we've never encountered before.
What's next for AI-Driven Recycle Decision Maker
Ultimately data is everything, if we were able to provide our model with more data and more time me and my team are sure we would have had a near perfect AI for the purposes of what we are using it for so in the future we hope to improve the model and tweak it as necessary. Additionally, we were thinking about adding computer vision instead of uploading the photo, however, that is a feature for the future that we were not able to implement this hackathon.
Log in or sign up for Devpost to join the conversation.