Seeking Guidance on Replacing 8th Wall Image Tracking with a Third-Party Solution

Technical Support
,
1

Greetings.

As 8th Wall’s image tracking will be discontinued this March—making it impossible to manage image targets—I’m exploring third-party alternatives to replace native 8th Wall image tracking.

I’m relatively new to 8th Wall and AR development in general, but I’ve managed to set up a self-hosted project using an app key, combining 8th Wall’s native image and world tracking as a starting point.

So far, I’ve tried using MindAR to replace the image tracking component, but I’m having difficulty integrating it with 8th Wall’s world tracking. My current goal is to recreate this experience where an image is detected, geometry is placed in front of it, and that geometry remains anchored in the world—even after the image target is lost—without drifting, sliding, or odd behavior, and while remaining observable from different angles.

Unfortunately, my attempts so far have been disappointing. If anyone has experience combining MindAR (or another image tracking library) with 8th Wall’s world tracking, I would greatly appreciate any advice or guidance.

Thank you in advance.

2

you can create handling with image tracking send event to spawn the model. use ar.js or mindAR send event and 8thwall get the event to show the model. i create like this for workaround

3

Thank you for response. I am trying to do it with MindAR now, though so far I am struggling even to pass anything from MindAR to an 8thwall in general and force model to stay even when target image is lost.

I am confused about multiple things:

  1. What should I do with camera feed? By default MindAR getting the camera and detects the image. Should we kill MindAR to free the camera and pass the event to spawn a model?

  2. What about position of the model that will be shown? Do we pass the offset from MindAR to 8thwall too?

  3. Do you move model from being a child of mind-ar-target to being a child of the a-scene?

It would be very great to see some examples for a reference if possible.

4

the FAQ states that image targets will be supported in the binary they distribute. Below is an excerpt. Am i missing something?

The distributed engine binary includes the core AR capabilities that power 8th Wall experiences, including:

  • World Effects

  • Face Effects

  • Image Targets

  • Sky Effects

  • Absolute Scale

5

Thanks for commenting. I hadn’t read through the FAQ thoroughly until now. Actually, using 8thwall wasn’t even my primary responsibility until recently.

It seems it will be closed-source binaries, but it should still be possible to self-host projects using 8thwall’s image targeting and SLAM combined. That’s a relief for me, but I’m not sure what my higher-ups think about that. They’ve tasked me with merging—what I consider for now to be unmergeable—an open-source image tracking solution with 8thwall’s world tracking. That’s why I’m looking for help here; perhaps somebody has done something like this before.