The idea of aframe-ai is to lower the entry-barrier for designers, students and the general public to explore ideas and create prototypes powered by AI.

A-Frame can act as an AI-Simulator like this:

1. Capture input:
   • capture images from the current VR-scene
   • capture images from the web cam looking at the user
   • capture audio from the user
   • capture audio from the current VR-scene
2. Send the input to a machine-learning service
   • e.g. RunwayML
   • e.g. DeepAI
   • e.g. OpenAI
3. Use the result to:
   • control non-human-players / robots
   • simulate assistive technology in the browser

This repo will showcase some examples of this approach...

• Assistive-Aframe — AI describing the virtual world to vision-impaired users
• Voice-Control — Use your voice to navigate a virtual world
• Head-Control — Move your head to navigate a virtual world