Future Work & Vision
Current Impact
What we developed in this project could actually change how visually impaired individuals interact with the world around them, especially in two specific and often challenging areas: using ATMs and navigating complex environments like campuses. Instead of needing outside help for basic but critical tasks like withdrawing money or walking across campus safely, our system gives them a real chance at doing it all by themselves. That's a pretty big deal.
ATM Module Achievements
The ATM module lets users place their finger on the screen and receive spoken feedback based on where their finger is located on the screen. Helping them navigate through ATM monitor with ease. Behind the scenes, this is powered by YOLOv8 for finger and button detection, and EasyOCR helps read the screen text.
Navigation Module Progress
The navigation module works similarly. We collected thousands of images from our campus and trained YOLOv8 to recognize common obstacles such as signs, poles, benches, and so on. The system can alert users with warnings if something dangerous is nearby. So that, they can avoid accidents and navigate with more confidence.
Technical Opportunities
Looking ahead, there's a lot more that can be done. More data would make the models more accurate, especially in new or changing environments. And using faster, more lightweight AI models would make this easier to run on smaller devices, like smartphones or even smart glasses. That's a big opportunity.
User-Centered Development
There's also the human side. If we had more feedback from visually impaired users themselves, we could make the system even more user-friendly. Maybe even test it in real-world places like a bank or a school and see how well it holds up.
In short, while this is just a concept for now, it shows real promise. With more time and support, we believe this could grow into something powerful, something that genuinely improves people's lives.