(Re)animating a Walmart skeleton with Raspberry Pi and computer vision.
- Can see with a camera
- Can hear with a microphone
- Can hear with a microphone array
- Can speak via a speaker
- Can rotate head
- Can rotate left arm
- Can rotate right arm
- Can identify presence and location of faces
- Can recognize faces from trained images
- Can follow faces with head rotation
- Can open/close mouth
- Can transcribe heard speech to text
- Can use speech to text
- Can interact with LLMs
ryan = Skeleton()
target = ryan.find_face()
if target is not None:
ryan.look_at(target)
ryan.say("Hello!")
prompt = ryan.listen()
response = ryan.process(prompt)
ryan.say(response)