About inference speed of Audio-Driven Facial Reenactment

[forest520]

Could you tell me what is the inference speed of Audio-Driven Facial Reenactment on Nvidia GPU？ Is it possible to execute in real-time such as 30fps? Thanks!

[RenYurui]

Hi, Thanks for the interest. We have not performed such estimations before. However, you can first try to estimate the speed for the image generation part using the provided code. It is better to modify the code to make different frames operate in parallel to increase the speed. Yurui

[liushengtong-nreal]

Hello, I have the same problem now, do you know the inference speed of Audio-Driven Facial Reenactment on Nvidia GPU now?Could you please share your conclusions, I would be grateful if you could.