We are obsessed with bringing self-improving visual intelligence to the physical world — the warehouses, forecourts, kitchens, and lines where the work actually happens.
Humans navigate the visual world quite robustly and perform many types of useful work using vision. It has been challenging to bring similar abilities to machines, but in the age of large vision language models, that is starting to become possible. OpenVector is exciting because it makes this technology work in real-time, for any camera, which can enable the automation of many types of labor.
Professor Shimon UllmanEECS, MIT
Affiliations & recognition
The launch
Watch the launch.
A two-minute look at what OpenVector does and why we built it.
