Smartphone-Captured Ear and Voice Database in Degraded Conditions
Sana Boujnah, Sami Jaballah, Mohamed Lassaad Ammari · 2018
The identification of persons through earprint and voice print is one of the basic biometric matters. The context of use is situated in smart homes. The main issue is the lack of databases containing voice records and ear images from the same persons collected with a smart phone. In this paper, we propose a new database for the persons identification through ear and voice print in degraded conditions. The database contains 111 subjects with more than 12 samples for each one. The new database is collected in degraded conditions, the ear images have different occlusion as scarf, hair, glasses, earrings while the voice records are collected with different types of noise in the background as music, cars, street, conversation. Interested researchers could contact the author to acquire the database.