The app takes an audio file as input and utilises  a CNN (Convolutional Neural Network) to determine whether the audio is synthetic or genuine. After classification, it uses GradCAM (Gradient-weighted Class Activation Mapping) to generate saliency maps, visually highlighting which parts of the audio contributed most to the model's decision. These saliency maps provide users with an interpretable explanation of the results, illustrating which audio features were most influential in the classification process. The app is designed to assist users in identifying manipulated or synthesized audio content, contributing to enhanced transparency and reliability in audio verification.