ai4r / Gesture-Generation-from-Trimodal-Context

Speech Gesture Generation from the Trimodal Context of Text, Audio, and Speaker Identity (SIGGRAPH Asia 2020)
Other
245 stars 35 forks source link

Code for preprocessing the data #18

Closed UttaranB127 closed 3 years ago

UttaranB127 commented 3 years ago

Hi, thanks a lot for sharing the preprocessed TED dataset with the repo! Will it be possible to share the preprocessing code as well, so I can use it for other datasets?

youngwoo-yoon commented 3 years ago

Sorry, the script is not available. Please see https://github.com/youngwoo-yoon/youtube-gesture-dataset. This is not the script used to make LMDB datasets, but mostly the same except two followings: 1) use of LMDB (for faster loading), 2) extracting 3D poses (by using https://github.com/facebookresearch/VideoPose3D).

UttaranB127 commented 3 years ago

I understand, thanks for the pointers!