cvlab-kaist / GaussianTalker

Official implementation of “GaussianTalker: Real-Time High-Fidelity Talking Head Synthesis with Audio-Driven 3D Gaussian Splatting” by Kyusun Cho, Joungbin Lee, Heeji Yoon, Yeobin Hong, Jaehoon Ko, Sangjun Ahn and Seungryong Kim
Other
295 stars 36 forks source link

The lip shape and audio of the trained video are out of sync. Do the training parameters need to be adjusted? #23

Open yzy1215-source opened 5 months ago