423
Lumiere’s language model allows it to interpret users’ textual instructions with high comprehension accuracy. It can animate static pictures, add or remove objects and characters from videos, and guess hidden parts of images. The AI can copy the styles or plot of the original content and create many different versions based on it.