In this video, I tried out various topics using video2video with anim diffusion. I used a multi control net for machine vision. The main emphasis was on length rather than consistency.