I just wanted to find out what happens if I always take the last frame of a video and use it as the input picture for the next 5sec video. And do this over 30+ times.
It's very interessting at the end the AI was not able to tell the content of the picture so he decied it is just a picture a girl is holding into the camera:D
I speed up the video in the middle to compress it from 3min to 1min runtime