Method and system for verifying dynamic handwriting and signatures by means of deep learning

A novel aspect is the inclusion of a first stage of temporary alignment of biometric characteristics, improving the training process of the deep learning based systems. Another novelty is the development of more robust biometrics systems. Finally, it does not require retraining of biometric systems when produces the registration of new users in the application. The method can be used both verifying signatures by means of deep learning The method can be used for both signature verification and handwritten password verification made using either a stylus or finger on a touchscreen. Therefore, it can be applied in the legal field as well as in the banking sector.