This shows you the differences between two versions of the page.

Link to this comparison view

caller_identification_by_voice [2014/03/15 12:21] (current)
bziolko created
Line 1: Line 1:
 +Magdalena Igras, Marcin Witkowski, Joanna Grzybowska, Paweł Jaciów, Mariusz Ziółko (Poland)
 +**Caller identification by voice**
 +The aim of our work is to develop the software for caller identification by voice. Based on collected speech samples, our system aims to identify emergency callers both online and offline. This homeland security project covers speaker recognition (if speaker’s speech sample is known), speaker’s gender and age detection and recognition of emotions. The propesed system is not limited to biometrics. The goal of this application is to provide an innovative tool for rapid and accurate threat detection and threat neutralization. This complex system will include: a speech signal analysis, an automatic development of speech patterns database and an appropriate classification method. 
 +Based on the preliminary data analysis using the proposed system, we concluded that a 20 sec long speech sample stored in the database is sufficient to identify a calling person. In other cases (i.e. absence of speech sample in the data base) a caller profile is defined by advanced signal processing techniques enabling efficient pattern matching. Automatic conclusions are based on speech signal comparisons. As a result, a caller is identified or his/her profile is described. The proposed system will be able to register and analyze acoustic effects that people may not notice in the recorded voice and in the acoustic background. The adaptive character of system aims at improving its efficiency.
Copyright © XXII PVC Organizing Committee 2013. All Rights Reserved.