Skip to content

visionOS 2 + Object Tracking + ARKit means: we can create visual highlights of real world objects around us and have those visualizations respond to the proximity of our hands.

License

Notifications You must be signed in to change notification settings

robomex/visionOS-2-Object-Tracking-Demo

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

9 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

visionOS 2 Object Tracking Demo

visionOS 2 + Object Tracking + ARKit means: we can create visual highlights of real world objects around us and have those visualizations respond to the proximity of our hands.

This project is largely a quick repurposing and combining of Apple's Scene Reconstruction sample project (which utilizes ARKit's HandTrackingProvider) and Object Tracking sample project.

The full demo video with sound is here.

Some details about putting together this demo are over here.

Build Instructions

  1. Choose your Apple Developer Account in: Signing & Capabilities
  2. Build

Models Used in This Project

I live in Chicago and purchased the cereal and milk at a local Jewel in June 2024 – your local packaging may vary and prevent recognition. The three products used are:

  1. Cap'n Crunch (Large Size)
  2. Fairlife 2%
  3. Lodge Dutch Oven

Using Your Own Models

If you want to strip out the three bundled objects and use your own:

  1. You will need to train on a .udsz file to create a .referenceObject, I recommend using Apple's Object Capture sample project to create a .usdz file of your object
  2. You will need to use Create ML (version 6, or higher, which comes bundled with Xcode 16) to train a .referenceObject from your .usdz, for me this process has taken anywhere from 4 - 16 hours per .referenceObject
  3. You will need to bundle your new .referenceObject in the Xcode project
  4. You will need to coordinate the naming of your new .referenceObject with the demo's ObjectType enum so everything plays nicely together

visionOS 2 Object Tracking Demo Clip