Inspiration
Our journey began with a profound empathy for the daily challenges of visually impaired individuals. We were driven by the belief that technology could make a meaningful difference in their lives. It is this passion for improving accessibility and quality of life that ignited the creation of ESN.
What it does
Our ESN project represents a transformative leap in assisting the visually impaired. We have developed a sophisticated neural network by fusing real-time Visual Simultaneous Localization and Mapping (vSLAM) techniques. With its intricate layer of nodes and weighted distribution, this neural network provides accurate proximity measurements of objects ranging from 0 to 1. Our solution empowers users by translating visual data into actionable, real-time information, enhancing their spatial awareness and independence.
How we built it
Our team harnesses advanced AI techniques, utilizing OpenCV and deep learning algorithms to create a unique application that assists the visually impaired. Real-time image analysis, distance calculations, and a neural network powered by Visual SLAM and OpenCV allow for accurate proximity measurements and drive a dynamic vibration feedback system.
Challenges we ran into
We faced a significant challenge: accurately distinguishing the floor and ceiling from recognized objects in our vSLAM and neural network system. This required fine-tuning our neural network to better understand spatial context and depth perception. Solving this issue was technically crucial for a user-friendly experience, as false positives could confuse visually impaired users.
Accomplishments that we're proud of
We're proud to have our application running and working!
What we learned
We learned that vSLAM systems are not easy to implement, as factors such as the camera and the general hardware can greatly impact results, and to keep it as stable as possible, we need to be consistent across the board.
What's next?
We hope to expand our products to everyone!

Log in or sign up for Devpost to join the conversation.