Translated data: Tencent has open-sourced the AniPortrait project, which generates high-quality animations using audio and facial photographs. This project extracts audio data, converts it into two-dimensional facial contours, and then utilizes diffusion models to create realistic and smooth facial animations. Experimental results show that AniPortrait has advantages in facial expression, pose diversity, and visual effects. Users can adjust facial movements according to their needs, enhancing practicality and fun. The project is directly open-sourced, providing convenience for users.