The AI model has a pose estimation component which synthesizes body figure images from video frames extracted from a training data set.

Udacity Researchers Develop AI to Generate Lecture Videos From Audio Narration

Creating content for MOOC platforms might not just be academically rewarding but consumes significant amount of time. On that account, Udacity research scientists have developed a machine learning framework that facilitates in automatically generating lecture videos from audio narration. The company also claims that their AI system can synthesize footage of any length by directly mapping between audio and corresponding visuals.