Avatar image animation using translation vectors

This invention describes a method to make an avatar image display emotions. It works by first training an artificial intelligence system to generate various emotional faces. From this training, the system learns specific internal representations, or "translation vectors," for different emotions. When a real person's face is analyzed to detect their current emotion, these learned emotional representations are applied to the avatar image. The result is an animated avatar that mirrors the individual's facial expression.