Meanwhile, OmniHuman-1 transforms photos into full-fledged videos
The highlight is that the neural network can produce quality lip-syncing and utilizes surrounding objects.
There is no code available, but you can check out other examples: omnihuman-lab.github.io