Lexi Learn

Lexi Learn vocab
Lexi Learn input
Lexi Learn Home

Inspiration

Who can best understand problems faced by students but students themselves? Lexi Learn has been built by four university students who are venturing into the world of programming and are trying to solve some major societal issues. With Lexi Learn, we aim to help anyone who wishes to enhance their English language comprehension. If someone is struggling with English for whatever reason -- English is not their primary language, they may be suffering from reading/learning disabilities like ADHD or dyslexia, or they are unable to grasp English even after putting in a lot of effort -- then Lexi Learn is the way to go! Coming from India - all four of us- know how big of an issue English-comprehension can be for non-native speakers.

What it does

As the World's first free large-text-to-storybook web application, Lexi Learn leverages cutting-edge generative AI to transform text into engaging visuals, enhancing comprehension & learning experiences effortlessly. All the user has to do is input any large PDF text file and click on generate -- our API framework uses generative AI to generate images for each portion of the text and enable visual comprehension. We aim to make the user understand what they couldn't understand earlier by outputting the text side-by-side with context-relevant images. We use learning-disability-friendly fonts and colors. We also have a vocabulary section where the user can store any words that they wish to learn the definitions and synonyms for. We have smartly used the technology of storing the data locally using cache so that they can refer to their vocabulary permanently without creating an account. We as students know how big a task it is to learn new words and how complex words can be a hindrance to understanding texts. Another cool feature of Lexi Learn is our "Quiz" section where you can test all your vocabulary and put your learning to a test multiple times.

How we built it

We have created a strong backend in Python and Flask with a frontend in HTML and CSS to create our open-access web application. We coordinated our files locally on VS-Code and finally uploaded the GitHub at the end of our project. We had to naturally remove our API key when we uploaded our code on a public GitHub repository.

Challenges we ran into

Our code is heavily dependent on prompt engineering to make sure that the AI generates the images of the input text as precisely as possible. If our prompt even goes wrong slightly, the AI response images will be biased and we had to make sure that DOES NOT happen. We kept at it and kept modifying our prompts. We had to test it out continuously with a bunch of different contextual texts. That took a lot of our time but we finally got the perfect AI responses thanks to our appropriate requests. While coding, our backend was stuck with a CORS error and we tried debugging it for a couple of hours.

Accomplishments that we're proud of

Coming up with a solution to this major academic issue in the span of less than 2 days has been a huge accomplishment for us. With students of all ages and backgrounds as our target audience, we the developers ourselves being students is a big deal because we can relate the best to them. We are actually really proud of the our possible scalabilities.

What we learned

We have created a strong backend in Python and Flask with a frontend in HTML and CSS to create our open-access web application. We were able to improve our understanding of these programming languages. We also learned multiple-API handling. And debugging of the CORS error which, according to what we have learnt, is a big programming issue.

What's next for Lexi Learn

In the future, we aim to further improve the inclusivity of our platform through text-to-speech generation and foreign language translation. We also aim to add another Quiz section where the user can test their understanding of their input text. We also aim to implement Machine Learning in this part and give the user recommendations of educational resources. Additionally, we aim to provide an age-appropriate text-summarisation for texts soon. Be on the lookout for our new features on our upcoming cool features that will only enhance user experience exponentially.