Hi everyone, I'm working with Core ML for object tracking and have successfully trained a couple of models. However, when I try to use the reference object file in the object tracking sample code from the WWDC video, it doesn't work. I'm training the model on a single-color plastic spherical object. Could this be the reason for the issue? I also attempted to use USDZ 3D assets that resemble the real object. Do these need to be trained with the Object Capture app to work properly? Speaking of the Object Capture app, my experience hasn't been great. When I scanned my spherical object, the result was far from a sphere—it looked more like a mushy dough. Any insights or suggestions would be greatly appreciated!

Hi, With this code on iPad Pro (11-inch) (4th generation), when capturing in area mode and pressing the "Next" Button, an overlay with a white dot in the center appears. Everything else is blurred out and the app is stuck. It does seem to work well on iPhones though. Thanks!

Hi, I am trying to train an object tracking model on MacBook Air with a M2 silicon. The .usdz file used to train is captured by iPhone 14 pro. However the training process takes too much time. it stay at 0.0% for about an hour and a half. I wonder if there is any other methods to generate reference objects.

Hello, I am trying to develop an app that broadcasts what the user sees via Apple Vision Pro. I have applied for and obtained the Enterprise API and actually can stream via the "Main camera access" API, as reported on https://developer.apple.com/videos/play/wwdc2024/10139/. My problem is that I have not found any reference to how to integrate the "Passthrough in screen capture" API into my project. Have any of you been able to do this? Thank you

I noticed that tracking moving images is super slow on visionOS. Although the anchor update closure is called multiple times per second, the anchor's transform seems to be updated only once in a while. Another issue might be that the SwiftUI isn't updating more often. On iOS, image tracking is pretty smooth. Is there a way to speed this up somehow on visionOS, too?

Hello, I have rendered an usdz File using sceneKit's .write() method on the displayed scene. Once I load it on another RealityKit's ARView using the .nonAR mode of the camera, I am trying to use the view's raycast(from:,allowing:,alignment:) method, to get the coordinates on the model. I have applied the collisionComponents when loading the model using the .generateCollisionShapes() function to be able to interact with the modelEntity. However, the raycast result returns nothing. Is there something I am missing for it to work? Thanks!

Anybody try hand tracking provider in 2.0? I'm getting them in 11ms interval, as advertised, but they are duplicate. Here's a print of the timestamps. Problematic for me because I am tracking the last 5 position for a calculation and expect them to be unique. Can't find docs on this anywhere. I understand it's not truly 90 updates a second but predicted pose, however I expected the updates to include predicted poses.

As the question suggests, I would like to use environmental awareness and item placement functions in Unity. Does have any related example projects?

When I wanted to call the Reality Composer Pro scene containing Object Tracking, I tried the following code: RealityView { content in if let model = try? await Entity(named: "Scene", in: realityKitContentBundle) { content.add(model) } } Obviously, this is wrong. We need to add some configurations that can enable Object Tracking to Reality View. What do we need to add? Note:I have seen https://developer.apple.com/videos/play/wwdc2024/10101/, but I don't know much about it.

The project was developed using Unity, and the requirement is to place a virtual model in the real world. When the user leaves the environment or the machine is turned off and then on again, the virtual model is still in its original real position. I found that the worldtracking function of Arkit is useful, but I don't know how to use it in Unity. Is that have any related example projects?

We are developing apps for visionOS and need the following capabilities for a consumer app: access to the main camera, to let users shoot photos and videos reading QR codes, to trigger the download of additional content So I was really happy when I noticed that visionOS 2.0 has these features. However, I was shocked when I also realized that these capabilities are restricted to enterprise customers only: https://developer.apple.com/videos/play/wwdc2024/10139/ I think that Apple is shooting itself into the foot with these restrictions. I can understand that privacy is important, but these limitations restrict potential use cases for this platform drastically, even in consumer space. IMHO Apple should decide if they want to target consumers in the first place, or if they want to go the Hololens / MagicLeap route and mainly satisfy enterprise customers and their respective devs. With the current setup, Apple is risking to push devs away to other platforms where they have more freedom to create great apps.

I tested the new visionOS object tracking and it worked really well. I have created a reference object using Create ML and it really detected the object. My question is: does it work also with iOS and, if not right now, is it planned to work in mobile iOS in the future?

hello i watched WWDC24, Ultra-Wide Mac Display. i wanna use to my player like that Ultra-wide mac Display. i wanna play for mp4 movie file in Ultra-wide mode (like that curved mode) Can i use Ultra-Wide AVKit, VisionOS 2 ? when i check in Apple documentation, AVExperienceController.Experience.expanded, Is this the function(Ultra-wide mode) I think it is? (https://developer.apple.com/documentation/avkit/avexperiencecontroller/experience-swift.enum/expanded#discussion)

When I wanted to call the Reality Composer Pro scene containing Object Tracking, I tried the following code: RealityView { content in if let model = try? await Entity(named: "Scene", in: realityKitContentBundle) { content.add(model) } } Obviously, this is wrong. We need to add some configurations that can enable Object Tracking to Reality View. What do we need to add?

It's a common system interaction to look at an item in SwiftUI and tap to select it. I'm confused how to do the same with ModelEntities. How do I use gaze to select a ModelEntity for context based actions? e.g. look at the green sphere and tap to pull up a menu. Or look in a direction and clap to **** away virtual objects etc. etc. If this is not possible is there a workaround?

Is it possible to determine where walls are in a shared space setting? Or does it have to be in immersive mode? Are there any workarounds for getting location of walls in shared settings? I want things to be able to latch onto walls.

I was following Explore object tracking for visionOS to load an object reference, but got this error: Failed to load reference object from URL: ObjectTrackingProvider.Error(code: referenceObjectLoadingFailed, errorDescription: "The operation couldn’t be completed. (com.apple.arkit error 1101.)", failureReason: "", recoverySuggestion: "" Here is what I have, not sure if it is an code error, or something with the system: private func loadReferenceObject() { Task { // Load the reference object let refObjURL = Bundle.main.url(forResource: "objectTrackerBox", withExtension: ".referenceobject") if let refObjURL = refObjURL { do { let refObj = try await ReferenceObject(from: refObjURL) logMessage = "Reference object loaded successfully: \(refObj)" print(logMessage) } catch { logMessage = "Failed to load reference object from URL: \(error)" print(logMessage) } } else { logMessage = "Failed to find the reference object file." print(logMessage) } } }

Hi, Object Capture's original sample code was released last year, and this year there was a talk about adding area mode to it. The talk links to the old Object Capture code - when can I expect to have the new one with area mode, and is there anything I can help you with to have it published faster? Thanks!

Hey, In the "Explore object tracking for visionOS" session we explore how a Globe can be tracked, and objects can be anchored to various positions. My question is if the physical Globe is rotated, will the anchored objects also respond to this in real-time? I would like to overlap a virtual map on top of a physical globe, so when the user rotates the physical globe, the virtual map also seamlessly responds. Is this possible using Object Tracking? Thanks

Are you planning on publishing a complete sample code project related to the Explore object tracking for visionOS session (wwdc2024/10101)? The animation at 12:50 where the globe opens up was especially impressive. Seeing how that was done while tracking to the globe would be very interesting. (I realize that we would have to create our own globe object in order for the code to work.)