AVmod is a Audiovisual modulator developed with Deep Fake

Face recognition uses a pre-trained model (keras-vggface) to embedding human face to compare (cosine distance) with the embedded face in the database

AVmod is a Audiovisual modulator developed with Deep Fake

My Project contains 8,664 photos featuring 100 distinct actors, Utilizing cosine similarity calculations, the model effectively identifies and retrieves the most akin images that bear resemblance to the inputted facial features, enabling precise face matching capabilities. I have used keras-vgg_face to efficiently extract the features of face.