Semantic Segmentation for a Coachella POC

Hello Lightship Voyagers!

I’m a senior engineer working at Trigger, a mixed reality agency. We’ve been super fortunate enough to have had access to the ARDK for a little while now and we’ve been actively using it across several projects and proof of concepts.

Today I’m going to be showing off how we leveraged the semantic segmentation feature of ARDK for a proof of concept we made for the Coachella music festival. The goal here was to specifically recognize the “Spectra” art installation at the festival grounds and draw Coachella branded 3D with animations on and around the art piece.

For this proof of concept we leveraged the layer mask features of semantic segmentation, combined with a custom skybox and some shaders to achieve the end product. We’re pretty satisfied with how it turned out and are really excited to revisit and continue iteration on this in the future!

Thanks again to Niantic and the Lightship program for making all of this possible!


Loved seeing this one in the launch demo! Nicely done :slight_smile:

1 Like

We show this off all the time to potential brand partners! Nicely done, trigger team!

1 Like