search

facebookresearch / multiface

Hosts the Multiface dataset, which is a multi-view dataset of multiple identities performing a sequence of facial expressions.
Other
718 stars 50 forks source link

Audio Alignment #15

Closed FreedomGu closed 1 year ago

FreedomGu commented 2 years ago

Hi, Thanks for sharing this fantastic dataset!

I was wondering if the audio is aligned with the images.

I downloaded audio and images and found some image parts of them are dropped.

Could you please tell me the frame rate you shot the videoes and how to align the data with audio by any chance?

songtoy commented 2 years ago

The frame rate should be 30, as mentioned in the paper. And I think the released data has been aligned with audio, though there are missing frames in some expressions.

alexanderrichard commented 1 year ago

The first available frame is aligned with the start of the audio file. You will have to account for dropped frames to keep audio-visual alignment in tact though.