Your problem, our scene
We look at what you are trying to detect and what your real data misses. If you have an Unreal scene, we work in it. If you do not, we say what building one costs before you commit to anything.
NameFrame is in Commercial Preview. That is not a marketing stage name: it means a small number of teams are using it for real work, with direct access to the people building it, and that we would rather do that well than sell it widely. If your perception problem is hard to collect data for, this is the way in.
Write to us about a pilotThis opens your mail client with the questions below already in the body. If you would rather not use email, everything we need is on this page: send it however you like to hello@getnameframe.com.
Not a trial licence and a link to the documentation. A pilot is us building the first dataset with you.
We look at what you are trying to detect and what your real data misses. If you have an Unreal scene, we work in it. If you do not, we say what building one costs before you commit to anything.
A few hundred frames, exported in your format, with the reports. You train on it. The point of a small run is that it either helps or it does not, and you find out quickly.
If it helped, we scale it: the same scenario, more frames, more conditions, automated through the API or MCP. If it did not, we would rather say so than sell you a subscription to a thing that is not working.
You do not have to talk to anyone to find out what the output looks like.
Twelve frames, labels, masks, depth, metadata and reports. Take the pack →
Distributions, gates and the check this capture failed. See the measurements →
The Studio, the CLI, the lenses, the weather, and how to get a first dataset out. Open the docs →