search

ShenhanQian / GaussianAvatars

[CVPR 2024 Highlight] The official repo for "GaussianAvatars: Photorealistic Head Avatars with Rigged 3D Gaussians"
https://shenhanqian.github.io/gaussian-avatars
Other
599 stars 92 forks source link

Length of Training data #27

Closed sstzal closed 6 months ago

sstzal commented 7 months ago

Hi， I have a question about the training data. Looking forward to your reply.

How many frames were used in your experiments during training a head avatar? I have conducted some experiments based on the original GS and found that using more frames for training leads to worse results. Do you have any suggestions as to why this is happening?

ShenhanQian commented 7 months ago

Hi, we use 1k-2k timesteps with 15 views to optimize an avatar.

According to my past experiments, more frames should contribute to better or at least comparable results unless face tracking is inaccurate.