Abstract: A mannequin comprising a head part and a torso part adjustably joined together. In the torso part a sound moving picture projector is mounted to project facial images on a film in a light beam upwardly through the neck into the head as the film is moved through the projector. Simultaneously sound from a track on the film is picked up, amplified and audibly reproduced. The sound and facial images are correlated to simulate a speaking person. The head part has a face piece with well defined forehead, nose, cheeks and chin but only smooth areas for eyes and mouth which is molded of transparent plastic and surface coated as a rear projection screen. Means in the head receive the light beam from the projector and transmit onto the rear projection screen so that the projected facial features correspond in location to the facial features on the face piece.