A modified TESPAR algorithm for wildlife sound classification

The application of sound classification for detecting intruders into protected wildlife regions has been addressed recently. For such an application, simple, yet robust, methods must be utilized since these systems usually are required to work in difficult conditions. The Time Encoded Signal Processing and Recognition (TESPAR) algorithm represents a good candidate due to its simplicity. In this paper we propose a modified TESPAR algorithm for wildlife sound classification. Comparative results obtained with the TESPAR algorithm and with the new modified version are shown. The new method has been tested on a database consisting of 300 sound recordings and several types of noise have been considered to simulate different possible environments.