search

facebookresearch / VisualVoice

Audio-Visual Speech Separation with Cross-Modal Consistency
Other
218 stars 35 forks source link

landmarks #7

Open hsl20130659 opened 3 years ago

hsl20130659 commented 3 years ago

Hi，thx for your great works， I am confused that which alignment algorithm you used and how many landmarks output ?

rhgao commented 3 years ago

The details are in Supp: http://vision.cs.utexas.edu/projects/VisualVoice/VisualVoice_Supp.pdf

rhgao commented 3 years ago

We use an SFD face detector to detect 68 facial landmarks.

debangliu commented 2 years ago

hi，thx for your great works，i confusing if the face is sideways, and can't be detected, what should we do