In the past few years, we have covered plenty of robots that can mimic human facial expressions. This video from Columbia Engineering shows an animatronic robotic face driven with learned models. Most current methods involved pre-programmed expressions but those are limited.
For this project, a robotic face with soft skin was used. It has a vision based self-supervised learning framework for facial mimicry. As the researchers explain:
By decomposing the learning process into a generative model and an inverse model, our framework can be trained using a single motor babbling dataset. Comprehensive evaluations show that our method enables accurate and diverse face mimicry across diverse human subjects.