Current implementation only successfully detects images generated via Style-GAN algorithm.
Just upload an image and then wait for the app to give it's prediction on whther the given face is a real one or generated by StyleGAN.
The app uses MediaPipe to identify regions of faces and send them one by one to the model to predict.
The model is based on a hybrid CNN model which uses ResNet-50 for Content Feature Extraction along with a Trace Feature Extractor.
