Digital human expression driving method and device, equipment and medium

• Main Authors: DAI CHANGJUN, FENG JINHENG
• Format: Patent
• Language: Chinese; English
• Published: 19.05.2023
• Subjects: CALCULATING; COMPUTING; COUNTING; ELECTRIC COMMUNICATION TECHNIQUE; ELECTRICITY; PHYSICS; PICTORIAL COMMUNICATION, e.g. TELEVISION

The invention relates to a digital human expression driving method and device, equipment and a medium in the field of network live broadcast, and the method comprises the steps: detecting whether a face image exists in a current image frame of a live broadcast video stream, and calling an expression driving model to generate a first face parameter of a digital human corresponding to the face image when the face image exists; detecting whether human voice information exists in the live video stream, and when the human voice information exists, calling a voice driving model to generate a second human face parameter of the digital human corresponding to the human voice information; smoothing the first face parameter and the second face parameter to obtain a target face parameter of the digital human; and constructing a face image of a digital human in the live video stream based on the target face parameter. According to the invention, expressions of the digital person and the person in the live video stream are