getting comparable results to commercial product #291

Open
opened 2022-06-30 10:11:40 +02:00 by tamirgold · 0 comments
tamirgold commented 2022-06-30 10:11:40 +02:00 (Migrated from github.com)

i played with the face swapper of a company called icons8 , it seems like they are using the same technique (probably the same code as well) and they are getting great results. besides training for a long time on the 512 VGGFace2 dataset what else can be done to get comparable results?

i played with the face swapper of a company called icons8 , it seems like they are using the same technique (probably the same code as well) and they are getting great results. besides training for a long time on the 512 VGGFace2 dataset what else can be done to get comparable results?
Sign in to join this conversation.