EcoEye AI

• 

  Our home page

• 

  Our AI page

Inspiration

Ever since we were children, we enjoyed playing outside and being outdoors, but as we grew up, we noticed more and more trash around the places we played. We went on garbage pick ups annually, but this did little to counteract the billions of people littering worldwide. Because of the problem pollution poses to our environment, we designed EcoEye. EcoEye is an AI that can detect trash in images.

What it does

Users can input an image and EcoEye AI will detect how many pieces of trash are in the given image.

How we built it

For the AI, we used YoloV5, a state of the art artificial intelligence model, to detect garbage in a given image. We used Flask to create a web server. Additionally, we used Vanilla JavaScript, HTML, and CSS for the front-end of the website. The front-end of the website communicates with the web server.

Challenges we ran into

Artificial intelligence is a nebulous concept and figuring out the best machine learning model to use was challenging. We had to go through multiple different iterations to figure out the best machine learning model to use, and finally, we decided to use YoloV5.

Accomplishments that we're proud of

We are proud of getting the artificial intelligence model working and getting a good final product. As our group consists of mostly beginner hackers, we were impressed that we had a working final product in such a short time frame.

What we learned

We learned more about transfer learning and using state of the art machine learning models in computer programs. We learned how to integrate the frontend of a website with the backend of a website.

What's next for EcoEye AI

Currently, EcoEye AI is hosted on a website where you can submit images, and it will identify trash in the images, but we hope to make it into a software that can be installed onto robots to turn them into automated trash cleaners.

Built With

• css3
• flask
• html5
• javascript
• pytorch