The Picsart Text2Video-Zero model leverages the power of existing text-to-image synthesis methods (e.g., Stable Diffusion), making them suitable for the video domain