spatialcomputing.info

https://preview.redd.it/mszr944xerb51.png?width=2048&format=png&auto=webp&s=f6f508539b444f6468643cdff38577fb22e724a8

Interface handles frontend and backend so that you can make a vision app with a single Python function.

As we move away from smartphone screens to augmented reality, we will be building and using apps that process context, like what you see and what you hear, rather than what you touch on screen.

I developed Interface to rapidly develop and ship my own ideas for AR. Shipping a new idea is as simple as:

1. writing one Python function that receives an image and returns an HTML (which can include text, images, videos, and any other custom interactions),
2. registering the function on Interface, and
3. sending a link associated with the function to users.

Check out this demo app to see what I mean: It simply returns a horizontally flipped grayscale version of the user image with arbitrary text in HTML. This proof of concept implies you can make any vision-based AR app today, such as overlaying ratings and reviews over dining menus.

Try It Yourself

Today is 8888th day since my birthday. To celebrate, I organized the project so that any developers can use Interface. Take a look at the Developer Documentation. Interface today only supports single instance photoshoots, and the documentation may be difficult to inexperienced developers, but I will soon provide guides with templates for inexperienced developers, improve user and developer experiences, and add platform capabilities such as sound and location to assist you in preparing for the era of AR glasses!

Disclaimer

Interface is my contribution, not a business. It receives no payments, includes no advertisements, and provides no service level agreement that guarantees the robustness you would want in a platform through which you would service production grade apps. I recommend that you use it for prototypes, not for products.