You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I’m currently working with the LivePortrait model for synchronizing expressions between videos and am seeking empirical advice on the best parameter settings for achieving accurate lip sync.
Specifically, I have been experimenting with the relative_motion parameter. When set to false, I observed excessive mouth closing, while setting it to true resulted in noticeable jittering.
Given that parameter settings can be complex and interdependent, I am looking for guidance on how to balance these parameters effectively. Could you provide any empirical advice or recommendations for setting parameters to achieve the best lip sync results?
Any insights or experiences you can share would be greatly appreciated.
Thank you!
The text was updated successfully, but these errors were encountered:
Hi @tonyabracadabra, you can try modifying this parameter. The Gradio interface allows you to directly adjust this parameter, which represents the strength of the driving force. You can tweak it until it meets your needs.
Hello,
I’m currently working with the LivePortrait model for synchronizing expressions between videos and am seeking empirical advice on the best parameter settings for achieving accurate lip sync.
Specifically, I have been experimenting with the relative_motion parameter. When set to false, I observed excessive mouth closing, while setting it to true resulted in noticeable jittering.
Given that parameter settings can be complex and interdependent, I am looking for guidance on how to balance these parameters effectively. Could you provide any empirical advice or recommendations for setting parameters to achieve the best lip sync results?
Any insights or experiences you can share would be greatly appreciated.
Thank you!
The text was updated successfully, but these errors were encountered: