search

OpenTalker / SadTalker

[CVPR 2023] SadTalker：Learning Realistic 3D Motion Coefficients for Stylized Audio-Driven Single Image Talking Face Animation
https://sadtalker.github.io/
Other
11.91k stars 2.22k forks source link

Major problem with Sad Talker #709

Closed Harsh-raj closed 11 months ago

Harsh-raj commented 11 months ago

Kudos to the creators, but I still can't figure it out why sadTalker was created with such an ancient version of pytorch, torchvision and torchaudio. This older version denies any attempt to utilise this model with any other existing technologies. If some solution to this problem is available please let me know. Also If there are alternatives to this model available that I am unaware of please let me know i would be greatful.

rachit2315 commented 11 months ago

There are a few open-source models that can be used for creating talking avatars similar to Sad Talker. Some of the most promising options include:

"DeepFaceLab" - an open-source tool for face swapping that can be used to create realistic talking avatars. "PoseNet" - a deep learning model that can detect human poses from images, which can be used to animate an avatar. "GANs" (Generative Adversarial Networks) - a type of AI model that can generate new images or videos based on a training set. This can be used to create talking avatars that look realistic and move in a natural way. These tools require some technical know-how to use, but they offer a lot of flexibility and can produce high-quality results.