Table 4

Comparison in terms of training time, no. of parameters and model size on DeepFor and VFHQ datasets for the case of 16 training videos.

MethodsF-SVMXceptDeFakeHopResNetLSTM
Number of parameters222.8M42.8K25.1M
Training time (DeepFor)0.7281s18m34s23.7251s1h15m24s
Training time (VFHQ)0.8584s17m45s27.9549s1h17m45s
Model size (DeepFor&VFHQ)10KB79.7MB76KB96.0MB
[ViewLarge]
This Feature Is Available To Subscribers Only

or Create an Account

Close Modal
Close Modal