TunnlVision

Tunnl Vision

Inspiration

In today's digital world, reading is more essential than ever. However, for individuals with dyslexia, accessing and comprehending text can be a significant challenge. While assistive tools exist, many lack flexibility and accessibility. Our application bridges this gap by leveraging web-based technology to provide a seamless, user-friendly solution for converting images of text into readable content.

What we've learned

Through this project, we enhanced our skills in structured file management and Firebase Cloud integration. Additionally, we explored AI-powered text recognition, enabling the application to accurately identify and interpret text within images while considering contextual relevance.

How we've built our project

We used Google Cloud to manage users' data on the application to handle certain things, to view images upon uploading, to enact transcriptions of those images, to provide AI-generated voice demos, and to view pictures afterward. We used an OCR (optical character recognition) to track the text to overlay them so that the readable parts still maintain the context in an image, making the image more readable.

Challenges faced

One of our biggest challenges was ensuring consistency in AI-driven text recognition. Variability in image formats and file handling posed difficulties, requiring extensive fine-tuning to optimize accuracy and performance. Another set of challenges was connecting the front and back end, which led to some bugs, mainly due to our splitting into different domains. Fortunately, we wanted to make code that's modular and versatile so that it won't be much of an issue, along with making multiple Git branches.