EyeQ

Ive been mainly working on bug fixes and just making everything a lot smoother, ive been testing with other smaller QOL features such as voice commands and being able to interact with the UI using your hands, Can’t wait to see where this goes!

The hackathon is over now, that was one of the craziest grinds ive done in a really long time, but now that it’s over, it was pretty fun. New features: Tutorial and you get to pick between featherless.ai and claude api, I tuned all the gesture controls up so they work nicer, more gestures like zooming, photo capture, stop ai response. my future plans is going to be trying to shrink it down and fitting all in a pi or something mobile so that you can decipher what your looking at anywhere, no need for a laptop or keyboard!

These past few hours have genuinely been crazy, my code stopped worked, worked, and then broke again. I got it fixed in the end but it wasn’t for nothing, I added a few features! I changed the Claude API for Featherless.AI models, I made it more gesture controlled (no keyboard required!) and more quality of life features such as smoothing, better gesture detection. I am far from done but I dont have much time so I can’t wait to see what I can get done! (its a lot harder to draw than it looks like) 

I started this project as a part of JecHacks Hackathon, where the theme was open ended innovation. This left me stumped for a while but I knew I wanted a project with Nvidia’s new LocateAnything model. So I laid the foundation for a software that allows you to ‘point’ at something and it can tell you whats inside. I plan on implementing the actual Nvidia model as currently its running on claude api, and more gesture controls for it to be fully keyboard free