Can a model be used on two different clips of the same person?
Hi,
I'm just trying to clarify another post on here, if you have a video clip of someone talking for ten minutes in reasonably static conditions eg. they are sat down at the same table in the same position and you faceswap the first minute of the video and make a model and alignments, can you not use that same information to convert faceswap minutes 1-2 without having to go through the entire training process again? Is this all down to frame by frame modelling and not really learning anything for future reference? Thanks.