Recognition of Individual Animals by their Vocalization

Kraljevski, Ivan and Grceva, Solza and Stojanovic, Igor and Gacovski, Zoran and Spireva, Biljana (2010) Recognition of Individual Animals by their Vocalization. In: the thirteenth Multiconference Information Society – IS 2010, 11-15 Oct 2010, Ljubljana, Slovenia.

[img] Text

Download (724Kb)


The recent advance in development of digital signal processing (DSP) and analysis techniques in human speech recognition systems has very significant influence on the field of bioacoustical research on animals [1]. In this paper, we present our approach in recognition of animal vocalizations with techniques for voice recognition. Vocalizations induced by physiological characteristics of the vocal tract of individual animals were encoded by Time Encoded Signals (TES). Further on, Artificial Neural Networks (ANN) was used for voice samples classification. The main benefit of the proposed technique is the application of such systems on low complexity DSP hardware due to low processing requirements [2]. In this paper, a system for individual “speaker” recognition of dogs barking samples was created and evaluated.

Item Type: Conference or Workshop Item (Paper)
Subjects: Natural sciences > Computer and information sciences
Divisions: Faculty of Computer Science
Depositing User: Igor Stojanovik
Date Deposited: 14 Nov 2012 13:12
Last Modified: 14 Nov 2012 13:12

Actions (login required)

View Item View Item