Open connormeaton opened 4 years ago
I am trying to implement this paper, which uses LipNet to create a 96d feature vector as features for training another model. Has anyone tried this? I want to pass a dataset of short videos and get features for each input.
I am trying to implement this paper, which uses LipNet to create a 96d feature vector as features for training another model. Has anyone tried this? I want to pass a dataset of short videos and get features for each input.