Inspiration
David and I both wanted to do an AI-based project but couldn’t find something that wasn’t too trivial but also wasn’t too hard. We finally found a project that wasn’t too hard but also wasn’t too easy: AI-powered handwriting detection. Because David was experienced in backend and AI development and I was more experienced with frontend and UI/IX design, we both thought that this project would be a perfect fit.
What it does
WAND recognizes English handwriting in any format, font, and text. It is trained to recognize words in the English alphabet that can be written on any surface and style, including computer text.
How we built it
David built the neural network and backend while I built the front end. David began by training the neural network in python while I began working on the frontend with html and css. I also created the link between the front end and backend with python. When we were done, David shared his code with me which I then merged with the backend. Following some further revisions to the front-end, WAND was finished.
Challenges we ran into
David and I both ran into significant challenges while building WAND. David had to train, design, and build the neural network which lead to heavy challenges as it was sort of charting through unknown waters. There were many errors which had to be solved via stack overflow. While building the front-end, I ran into many errors involving formatting and design. There were many times in which the animations or text was messed up and there were many errors which I had never seen before.
Accomplishments that we're proud of
We are proud of the relaxing UI/IX as well as the successful bug-less flow of the website. Users can successfully upload and see the results of their handwriting/ other images.
What we learned
David learned many new AI techniques and libraries while I learned more front-end styles and tools and a lot of existing CS concepts for us was reinforced.
Log in or sign up for Devpost to join the conversation.