It's very easy to add people to this video. There are two ways - a video transition from frame to frame and a text task. And a better one is a text task for an image (render). In this case, everything turns out to be clearer and without loss of detail and texture, movement always blurs the image, and static leaves everything as it was. Then you need to overlay these video clips with people on top of the finished video, add fade-outs and varying degrees of transparency around the edges so that people disappear and do not appear in the next video fragment. Everything is simple. See how it can be on the example of the first video. I added some people there. This is all with the help of text tasks. Of course, different AI models have their own peculiarities - some are more animated, some are more realistic. For example, they are more realistic in Kling 2.1 PRO, but as time goes on, everything changes.
Videos of people in the room 