*Consumers are finding it increasingly difficult to distinguish between healthy and unhealthy products as well as products that are good or bad for the environment. Plastic packaging from supermarkets alone have left Britain with 59 billion pieces. In addition to this, 3.6 million tonnes of food are wasted a year within the same nation. These are humungous statistics, and we plan to change this by educating users on the healthiness and environmental impact of the purchases they are about to buy. *
What is Shop Smart
Shop Smart is a new web-based concept that allows users to upload an image of their product which is then identified using AI that will also tell the user the given categories the item falls into.
We used HTML and CSS to develop our user-friendly website here is a snap shot of one of the pages.
A python library, known as PyTesseract has been used to create this model. PyTesseract allows the model to read the image and provide a full alphanumeric character recognition of the image. The library enables the extraction of text from the image and allows it to store in a string.
Secondly, the model runs a couple of functions, where the extracted text is compared against a huge dictionary containing specific words which, at the end give a Healthy Score and an Unhealthy Score. The model also prints a mean score and if the mean score is more than the sum of the constituents, we recognize it as sustainable.
To use our Python model with the API, which is based on HTML and CSS, we used a framework, called Flask, which can be used by importing it as a library in our Python Code. Flask allows coders to use GET and POST requests which allowed us to get the fetch of the upload image from the HTML form to a directory, which PyTesseract can read and evaluate results based on, and finally, we printed our results on a separate HTML page.
Mock Up App
Within the limited time, we thought it would be extremely difficult to create an app given our team's abilities so instead of this, we decided to create a mock-up version because we believe that this concept would be most successful when implemented into an app. Here are some pictures of the mock-up we created using Figma. If you would like to try it out, navigate to the link below!
Why are we sustainable?
We are sustainable because we are able to prevent customers from buying harmful products to the environment before purchasing. We hope to simultaneously prevent food waste by teaching users to eat healthily as well as decreasing unsustainable foods which will lower demand and therefore supply which reduces wasteful packing and large carbon footprints.
Why pick us?
We were able to execute a highly complex project that incorporated ML, where our group's abilities are to a high school level. There are very few products like ours on the market that are beneficial to society. Our idea is unique and has a lot of potential to do good in our world and fix some of the problems that society has made.
Our Key Takeaways
Accomplishments that we're proud of
We are extremely happy to have undertaken a very complex project in under a few hours as high school students. We worked really hard and communicated well, which allowed us to integrate machine learning into our product and well-designed user interfaces.
What we learned / challenges we faced
We learned to communicate effectively and be clear to each other in what we are saying. Initially, we did find coming up with a project quite hard as the criteria is so broad and we had no clue where to start but, luckily with a few minutes of brainstorming we were able to come up with our idea!
What's next for Shop Smart
In the future, we hope to code an app version of the website that allows people to scan their images as well as uploading them for user usability.