In my previous post, I documented how you can train a Machine Learning (ML) model that will enable the Vector to detect another Vector.

I am happy to announce that I am releasing my version of the Vector SDK that has inbuilt support to allow you to see Vector’s live camera feed with another Vector detected in it. Take a look at this video:

The details of how you can do the same are on my git repository. All you need in a pair of Vector robots. If you have a Vector space, that can come handy to ensure that the otehr Vector doesn’t get too far off. Really hope some of you can try this, because the next step would be to enable you to contribute to the public version of the Anki Vector dataset so that we can collectively make ML model richer and better. Once we get to a satisfactory level of performance, and hoping that Digital Dream Labs (DDL) delivers the complete OSKR, we could embed the ML model in Vector’s firmware and enable Vector to detect another Vector natively.

Please give it a try! And please let me know if you face obstacles, or any other feedback.