Path Finder

test case 1
test case 2
test case 3

Inspiration

We used the following questions to drive us through the lengthy, yet exciting hackathon experience: what does our school currently offer? How can those opportunities be improved? Using that as inspiration, we realized how much students worldwide were struggling to recover mentally, emotionally, and academically from the Covid-19 pandemic. Participating in PeddieHacks was an eye-opening experience to the tangible impacts high school and college students can make on their peers. By creating an OCR model we were able to provide a free learning platform to solve challenging problems, that we once struggled with in our math classes, and find more practice with one click.

What It Does

Students often struggle in math classes because of content gaps. Path Finder is a website that helps students find their way through math problems, identify any weaknesses, and helps them eliminate them for free. For example, if a student does not understand a tricky homework problem, instead of giving up, they can press the 'Struggling' button on our website, upload an image of their math problem, and get instant feedback. Our website will provide them with an answer and helpful resources, such as Khan Academy videos, that they can use to learn the topic in the future. Once they have achieved mastery of a type of question, they can continue by pressing the 'Practice' button, uploading the same picture for similar results. In the future, we strive to generate practice problems for the user to solve and help them prepare for their upcoming math exams. Our future goal is to combine OCR models and Machine Learning to provide a fast and effortless UI/UX system that strengthens a student's learning experience and is accessible to all.

How We Built It

Our website was mainly built using HTML, CSS, and JS frameworks. The code for the image-to-text algorithm was built using Pytesseract, an OCR tool for Python.

Challenges We Ran Into

The biggest challenge was creating the algorithm used to extract text from an image. We had attempted to use several different APIs, but none of them worked. There was either something wrong with the code or we had trouble downloading the necessary packages. We also had trouble directly implementing the Jupyter notebook file onto our website, so we ended up using PHP and JS.

Accomplishments that We're Proud of

Working with an OCR python model for the first time and building a beautiful, fully functional website!

What We Learned

Python OCR, implementing it through PHP and JS in a website, and UI/UX design.

What's next for Path Finder?

Chemistry, physics, and advanced math topics. For example, trigonometric functions and calculus. The function can also look for keywords (repetitions) in history passages to give similar resources about a topic.