How to Reengineer Your Computer Vision Models to Work on an Aftermarket Device

You said earlier that your perception stack or your original algorithms are able to run on sort of aftermarket devices. I'm assuming these are like your typical Android phones, which have a fairly powerful camera and then some amount of processing or compute on the device itself. Talk to me about why you thought we should take this route. And then once you've decided that we want our computer vision models to work on this kind of device, how did you go about reengineering it or engineering it in a way that they were able to do that?

Transcript

Play full episode

The AI-powered Podcast Player

Save insights by tapping your headphones, chat with episodes, discover the best highlights - and more!

Get the app