Asking from a technical standpoint because I feel like the term is doing a lot of work in coverage of this space right now. Genuine real-time video inference, where a model is generating or transforming frames continuously in response to a live input stream, is a fundamentally different problem from
