Skip to content
VISION Docs

Object Detection

Object Detection lets you ask VISION what’s in front of you. It captures the scene, analyzes it with AI, and speaks back what it recognizes.

How to Use

Hold the button and say any of these:

  • “Describe objects”
  • “What do you see”
  • “Look around”
  • “What can you see”
  • “What is ahead” / “What’s ahead”
  • “What do you detect”

Chinese: “识别物体”, “你看到什么”, “看看周围”, “前方有什么”

What to Expect

VISION describes the objects it recognizes in natural language, mentioning the most important ones first.

Example

You: “What do you see?”

VISION: “I see a person ahead, a bench to your left, and a car further away.”

How It Works (In Simple Terms)

  1. A camera on your glasses captures the scene in front of you
  2. An on-device AI recognizes objects (people, vehicles, furniture, and many more)
  3. The most important items are announced first so you hear what matters most

Object Detection is on-demand — it only runs when you ask, so it doesn’t drain the battery.

What VISION Recognizes

VISION can identify a wide variety of everyday objects, including:

  • People and common animals
  • Vehicles — cars, bikes, buses, trucks
  • Furniture — chairs, tables, beds, sofas
  • Food and drink items
  • Electronics — phones, laptops, screens
  • Household items

Priority

Not everything gets announced with equal importance. VISION prioritizes safety-critical objects:

  1. Highest priority — vehicles
  2. High priority — people
  3. Medium — most everyday objects
  4. Lower — small items like cups or books

This way you hear the most important information first.