Foreign accent recognition project (FARP)
Ville Hautamäki · 2015
If a traveler or immigrant has biometric passport, then verifying the validity of the passport can be done by measuring face, fingerprint or voice and matching it against the stored model in the chip in the passport. However, many countries do not have not implemented yet biometric passport technology, in addition some Finnish passport categories, such as temporary passport do not have the chip. In this project we study a form a form of softbiometric, which does not identify an individual but a group of individuals. In this work we concentrated on the country origin, that is a required field in all passports. If we detect the country of origin to be false then we say that possibly the passport is fake. We implement the country of origin detection by the way of detecting the English foreign accent. We have recently proposed a universal acoustic characterisation to foreign accent recognition, in which any spoken foreign accent was described in terms of a common set of fundamental speech attributes. Although experimental evidence demonstrated the feasibility of our approach, we believe that speech attributes, namely manner and place of articulation, can be better modelled by a deep neural network. In this work, we propose the use of deep neural network trained on telephone bandwidth material from different languages to improve the proposed universal acoustic characterisation.