We're developing a VisionOS application, where we would like to do product recognition (like food items). We have enterprise entitlements and therefore also main camera access for VisionOS. We send this live camera frames to a trained CoreML model where we will receive 2D coordinates from the model detection prediction. Now, we would like to create a 3D anchor on the detected items so it can be visible for user. The 3D anchor is going to be the class name of the detected item. How do we transform this 2D coordinate from the model prediction to a 3D anchor?

I'm using Unity 2022.3.56f, with Apple VisionOS App Mode set to 'Virtual Reality - Fully Immersive Space'. It seems that the render resolution of my game in the Apple Vision Pro when I build is well below the native resolution of the AVP displays. I can't see a setting in XR Plug-in Management Apple visionOS options, or in Quality settings, to increase the render resolutions. Is this possible? I tried setting: UnityEngine.XR.XRSettings.eyeTextureResolutionScale= 2.0f For example, but this doesn't seem to do anything to the render resolution in the build.

Hi, I'm developing an app for the Apple Vision Pro. Inside the app the user should be able to load objects from the web into the scene and then be able to move them around (dragging and rotating) via gestures. My question: I'm working with RealityKit and use RealityView.. I have no issues loading in one object and making it interactive by adding gestures to the entire RealityView via the .gestures() function. Also I succeeded in loading multiple objects into the scene. My problem is that I can't figure out how to add my gestures to multiple objects independently. I can't use Reality composer since I'm loading the objects dynamically into the scene. Using .gestures() doesn't work for multiple objects since every gesture needs to be targeted to a specific entity but I have multiple entities. I also tried defining a GestureComponent and adding it to every newly loaded entity but that doesn't seem to work as nothing happens, even though my gesture is targeted to every entity having my GestureComponent. I only found solutions that are not usable on Visionos / RealityView like installGestures. Also I tried following this guide: https://developer.apple.com/documentation/realitykit/transforming-realitykit-entities-with-gestures But I feel like there are things missing and contradictory, like a extension of RealityView is mentioned but not shown and the guide states that we don't need to store the translation values for every entity and therefore creates a EntityGestureState.swift file to store these values but then it's not used and a GestureStateComponent is used instead that was never mentioned and contradicts what was just said about not storing values per entity because a new instance of it is created for every entity. Nevermind, following the guide didn't lead me to a working solution.

Using Unity to develop VisionOS program, pressing the right knob of VisionPro during use will exit the Unity space and destroy the model in the space. The model in the space has been disconnected from the SwiftUI interface. After clicking the right knob, return to the system main interface, and then click the right knob again to return to the inside of the program. However, Unity space cannot be restored, and calling the discisWindow method on the SwiftUI interface has no effect, so the interface cannot be destroyed. Is there any solution??

Hi, I'm trying to achieve 3D photo effects using a photo and a depth map, using GeomertyModifier. The offset is getting applied correctly in Reality Composer Pro, and in Xcode, but when I launch the app, the model offset is not getting applied. here are the screenshots of shader graphs and how it looks in the app

Hello Community, I am currently developing an experimental VisionOS app, to investigate the social effects of the new Spatial Persona feature, for my bachelor thesis. My setup includes a simple board game for the participants, in which they can engage with their persona avatars. I tried to use the TabletopKit for this setup, but ran into issues when starting the SharePlay session. When I testes my app, I couldn't see the other spatial persona anymore, despite the green SharePlay button indicating the session started. The other person can see my actions in their version of the app on the board, but can not interact with anything. Also, we are both seat on the default side of the seat. I tried to remove the environment I added, because it doesn't seem to synch with the other player. When I tried the FaceTime feature in the simulator without the environment, I could then see the test robot avatar, but at a totally wrong place. It's seems like it isn't just my environment occluding the seats, but a flaw in the seating process as well. When I tried the FaceTime feature in the simulator on the official test scene (TabletopKit Sample), I got the same incorrect placement and the warning "role(for:inSeatNumber:): The provided role identifier does not match a role in the current template." So my questions are: What needs to be changed so the TabletopKit can handle seating correctly? How can I correctly use immersive scenes in combination with the TabletopKit? I tried to keep the implementation of the TabletopKit example as close as possible, so I think it will enough to look into this codebase for now. I debugged the position of seats and they are placed correctly in front of their equipment. The personas are just not placed on them.

I am experimenting with RealityKit to set up a portal. Everything works, but I was wondering where the scene's origin is with respect to the front of the portal window? From experiments, the origin's X and Y appear to be at the center of the portal window, while the origin's Z appearing to be about a meter behind the portal window. Is this (at least roughly) correct? Is it documented anywhere? PS. I began with the standard visionOS app and edited the Reality Composer Pro file to create the scene.

Still don't understand why no one is clarifying about this Apple Video https://developer.apple.com/videos/play/wwdc2023/10111 At the end of this video, there's an incomplete tutorial about connecting a USDZ with mesh and Skeleton structure to the hand tracking system. No example project is linked, and no one is giving the community any clarification. Please can you help us to understand how to proceed?

Hello, We discovered that a bunch of our old animated models were no longer animated on iOS15 and onwards. After a few days of playing spot the difference between usda files I noticed that all the broken models had an xform called "Scene". Lo and behold, changing the name of that xform fixed the issue on all the models. Even lowercase "scene" makes the animations work again. Is "Scene" a reserved keyword or something? What other keywords do we need to avoid so we can create more robust USDZ files? I'm surprised this issue isn't more widespread considering Blender wraps models in a "Scene" node. At the drive link below you can find two animated cube USDZs. The only difference is the name of one of the xforms. The one with a "Scene" xform is not animated in quicklook (replicated on iPhone 13 iOS v15.2, iPhone 13 iOS v 18.3, and various devices on Browserstack including iPhone 16 iOS v18.3). https://drive.google.com/drive/folders/1dch1WaM9O6mbHy29S6NGWgnSHkZkPiBf?usp=sharing

Hi ! I'm new on this forum, so if I need to update this post to have more info, or anything else, please let me know. I'm using the Apple Vision Pro to develop some app (with unity). To demonstrate what the user see on the headset, I would like to mirror the view on a device (an iPad in this case). I managed to do this without any issue. My problem is that, in the Vision Pro, I have an interface that the user can interact with. But I would like to be able to manage myself the interface on the iPad. What I mean is that the user can (or can't, doesn't matter) see the interface in the headset, and the interface is controlled by myself on the iPad. Is there any way to do this ? Is this a question I should ask on unity's forum ? (I don't think so, because it should be related to the mirroring function non ?)

Here is my code in visionOS 2.3 NavigationSplitView { List { } .navigationTitle("Passwords") } detail: { Text("Hello") .navigationTitle("All") } The font size of "Passwords" and "All" are smaller than the ones in Passwords app.

As I understand it there are two ways I can track a hand, or a joint, in RealityKit: either, create an AnchorEntity, for example AnchorEntity(.hand(.left, location: .palm)) or, set up an ARSession with a HandTrackingProvider ( a lot more code which I haven't repeated here). Assuming this is correct, when would I want to use one over the other?

I have a grpc server running inside of a task. When the user takes the headset off, the grpc server will no longer work when they put the headset back on. I would like to have this action detected so that I can cancel the task (which will effectively close the grpc server). I am also using a visual indicator to let the user know if the server is running, but it will not accurately reflect the state of the server when removing and putting back on the headset.

I would like to know if there have been any updates that improved the image tracking speed. I submitted feedback a while ago, mentioning that the image tracking rate was only one frame per second on Vision Pro. Thank you very much. FB15688516 and FB16745373is feedback we previously submitted, but we have not received any response yet.

I am using ARKit to detect image in visionPro. However I met some question about adding the reference image. Some of my images can not be added correctly sometimes. (As you can see in the picture above, the 'orange' can not be added correctly, but the 'cup' can). However, sometimes they will be added without any problem. I do not know why it will happen. And I want they all be added steadily.

We are using the ARKit image tracking feature on visionOS 2.0 with three pre-registered images. The image tracking works, but only one image is actively tracked at a time. When more than one target image is visible to the camera, it has difficulty detecting and tracking the other images. Is this the expected behavior in visionOS, or is there something we need to do to resolve this issue?

I am wondering, is it possible to somehow configure a 3D object to respond to the gaze of a person, like change colors of some parts of the 3D-Model where a person is looking, i.e. where a person's gaze lands on the surface of the 3D-model ? For example, if there is a 3D model of a Cool Dragon 🐉 in the physical space of a person, when seen through with the mixed reality view, of a VisionPro. Now, it would be really cool to change only the color, or make some specific parts of the dragon skin shimmer, but only in the areas where a person is looking. Is this possible ? Is it do-able with eye-tracking of VisionPro ? Any advice would be appreciated. 🤝 🤓 I am new to iOS and VisionOS development.

When I get close to an Entity in RealityKit wearing VisionPro. The Entity will become transparent so I can distinguish it is rendering by VisionPro instead of an object in reality world. How can I make it not transparent when I get close to the Entity?

Hello, we have a RealityKit app that also runs on macOS via Catalyst. For specific USD assets containing particle systems we have observed a reproducible crash. Steps to reproduce: Open Reality Composer Pro Create new file Create simple particle system (default one is fine) export as USDZ Create project in Xcode Call Entity.load(… and pass in your USD Running this on an Intel iMac with macOS Sequoia 15.3 will lead to a crash with the following console log: validateWithDevice:4704: failed assertion `Render Pipeline DescrvalidateWithDevice:4704: failed assertion `Render Pipeline Descriptor Validation depthAttachmentPixelFormat (MTLPixelFormatDepthvalidateWithDevice:4704: failed assertion `Render Pipeline Descriptor Validation depthAttachmentPixelFormat (MTLPixelFormatDepthvalidateWithDevice:4704: failed assertion `Render Pipeline Descriptor Validation depthAttachmentPixelFormat (MTLPixelFormatDepth32Float) and stencilAttachmentPixelFormat (MTLPixelFormatStencil8) must match. ' iptor Validation depthAttachmentPixelFormat (MTLPixelFormatDepth32Float) and stencilAttachmentPixelFormat (MTLPixelFormatStencil8) must match. ' 32Float) and stencilAttachmentPixelFormat (MTLPixelFormatStencil32Float) and stencilAttachmentPixelFormat (MTLPixelFormatStencil8) must match. ' 8) must match. ' Xcode version: 16.2.0 iMac 2020 3,8GHz Intel Core i7 macOS Sequoia 15.3 FB16477373 It would be great if this could be fixed quickly or a workaround provided since it affects or production app. Thank you!

Hello, I am developing a visionOS application and am interested in obtaining detailed data of users’ hands through ARKit, including but not limited to Transform and rotation angle. I have reviewed Happy Beem, but it appears to only introduce the method of identifying the user’s specific gestures. Could you please advise on how to obtain the Transform and rotation angle of the user’s hand? Thank you.