Visual Insight AI

front
esp32-cam
side

Inspiration

We set out to make AI feel less like something distant and more like a personal, everyday tool that blends into your life. Our goal was to make AI so seamless, it feels like a natural extension of your own perception. We've taken massive steps towards breaking down the barriers and layers between AI and you with these glasses.

What it does

These smart glasses bring AI closer than ever, offering insights and information without the usual delays or interruptions. Whether you're exploring new places, reading signs in foreign languages, or identifying objects around you, the glasses streamline the whole process making AI feel more immediate and personal.

How we built it

We crafted these glasses using Azure OpenAI and Azure AI Vision for interpreting your surroundings. OCR handles text detection, while Flask runs our backend, connecting all the pieces. To ensure smooth interaction, we used React and Framer Motion for an intuitive user interface. Although real-time functionality was a challenge, we’ve significantly minimized delays.

Challenges we ran into

Getting all the different technologies AI, OCR, vision processing, and the user interface connected and working harmoniously was no easy task on the software side of things. Balancing the trade-offs between resolution, refresh rate and speed on the hardware side was just as challenging. Each component had to communicate seamlessly, without breaking the flow of information. One of the biggest hurdles we faced was managing API rate limits while maintaining speed and accuracy. It required meticulous effort to keep the system running smoothly, making sure we didn't compromise on performance while navigating these constraints.

Accomplishments that we're proud of

This being our first hackathon, getting to complete a project together has been such an exciting and rewarding experience for all four of us. It’s been so much fun diving into new challenges and learning along the way. The opportunity to collaborate, experiment, and create something tangible has left us all feeling proud and motivated to take on even more in the future!

What we learned

We learned how critical optimization is when dealing with real-time data. Managing data flow between the glasses and the backend taught us a lot about performance. We also gained deep knowledge of integrating multiple AI technologies into a single user-friendly platform.

What's next for Visual Insight AI

We’re excited to take Visual Insight AI to the next level by improving voice interaction, adding more customization, and expanding its capabilities in more environments.